90132 The Business Models of mLabs and mHubs— An Evaluation of infoDev’s Mobile Innovation Support Pilots 2 2 The Business Models of mLabs and mHubs— An Evaluation of infoDev’s Mobile Innovation Support Pilots 2 3 The report “The Business Models of mLabs and mHubs—An Evaluation of infoDev’s Mobile Innovation Support Pilots” is available at https://www.infodev.org/mobilebusinessmodels Copyright © 2014 International Bank for Reconstruction and Development / The World Bank Mailing Address: MSN I9-900 1818 H St. NW, Washington D.C., 20433 USA Telephone: (+1) 202- 458-4070 Website: www.infoDev.org Email: info@infodev.org Twitter: @infoDev Facebook: /infoDevWBG Some rights reserved. This work is a product of the staff of infoDev / World Bank. Note that the World Bank does not necessarily own each component of the content included in the work. The World Bank therefore does not warrant that the use of the content contained in the work will not infringe on the rights of third parties. The risk of claims resulting from such infringement rests solely with you. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of the donors of infoDev, The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 Unported license (CC BY 3.0) http://creativecommons.org/licenses/by/3.0. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution – Please cite the work as follows: infoDev. 2014. The Business Models of mLabs and mHubs — An Evaluation of infoDev’s Mobile Innovation Support Pilots Washington, DC: World Bank. License: Creative Commons Attribution CC BY 3.0 Translations – If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. All queries on rights and licenses should be addressed to infoDev, The World Bank, MSN: I9-900, 1818 H Street NW, Washington, DC 20433, USA ; email: info@infodev.org Photo Credits infoDev / World Bank (photos on pages iv, 6, 10, 12, 16, 17, 23, 29, 44, 49); IFC (page 24); Jonathan Kalan (pages 32, 39, 41); iHub (pages 8, 42); mLab ECA (page 17); Zena Fruits (page 34); Caribbean Growth Forum / Digital Jam 2.0 (page 47). 2 4 Cover design: infoDev and Jean Manuel Wegimont About infoDev infoDev, a global trust fund program in the Financial and Private Sector Development Network of the World Bank Group, supports growth-oriented entrepreneurs through creative and path-break- ing venture enablers. It assists entrepreneurs to secure appro- priate early-stage financing; convening entrepreneurs, investors, policymakers, mentors, and other stakeholders for dialogue and action. We also produce cutting-edge knowledge products, closely linked to our work on the ground. About infoDev’s Mobile Innovation Program infoDev’s Mobile Innovation Program supports growth-oriented mobile app businesses in emerging and frontier markets. infoDev does this by 1) enabling entrepreneurship in the mobile industry, through venture incubation and acceleration, 2) building mobile innovation communities of entrepreneurs, investors, partners, and mentors, and 3) researching the app economy of emerging and frontier markets. The backbone of infoDev’s Mobile Innovation Program is a global network of Mobile Application Laboratories (mLabs) and Mobile Social Networking Hubs (mHubs), rolled out across eleven countries. 2 5 Acknowledgements T his report, commissioned by infoDev, a global partnership program within the World Bank, evaluates the business models of infoDev’s Mobile Application Laboratories (mLabs) and Mobile Social Networking Hubs (mHubs). mLabs and mHubs were implemented under the Creating Sustainable Businesses in the Knowledge Economy (CSBKE) program, funded by the government of Finland, in partnership with Nokia. Nicolas Friederici (Oxford Internet Institute) wrote the report. Toni Eliasz was task manager for infoDev. The task team leaders were Ellen Olafsen and Tim Kelly. The team is indebted to the more than 200 stakeholders that provided input and feedback for the analysis through interviews and focus groups. Special thanks are due to mLab and mHub managers, who helped with extensive results reporting and interviews, local facilitation during country visits, as well as reviewing. Several infoDev staff gave detailed comments during an internal review: Angela Bekkers, Brett Dickstein, Rick Doerr, Diletta Doretti, Carolyn Florey, and Zoe Lu. Colin Blackman copy-edited the report and Jean Manuel Wegimont designed the layout. The review and decision meeting was chaired by Chris Vein (Chief Innovation Officer for Global ICT Development at the World Bank). Peer reviewers were Chunlin Zhang (Lead Private Sector Development Specialist, World Bank), Mavis Ampah (Regional Coordinator for Sub-Saharan Africa, ICT Sector Unit, World Bank), Jussi Hinkkanen (Vice President for Corporate Relations and Business Environment for Nokia Middle-East and Africa), David Taverner and Marissa Drouillard (Mobile for Development, GSMA), and Tayo Akinyemi (Director, AfriLabs). The report was made possible by the support of the Ministry for Foreign Affairs of the Government of Finland. 6 2 Contents Executive Summary....................................................................................................................... 10 Part I: Rationale for a Business Model Evaluation of mLabs and mHubs..................................... 16 Introduction and Background................................................................................................. 16 Why Evaluate the Business Models of mLabs and mHubs?............................................ 25 Results for mLabs and mHubs.............................................................................................. 28 Part II: Lessons Learned and Future Directions for mLab and mHub Business Models.............................................................................................................................................. 32 Lessons Learned for mLabs and mHubs............................................................................ 32 Room for Debate: Open Questions for mLabs.................................................................... 45 Lessons Learned and Future Directions for infoDev........................................................ 53 Conclusion................................................................................................................................. 67 Part III: Case Studies of mLab and mHub Business Models.............................................................70 Applying the Business Model Canvas to mLabs and mHubs.......................................... 70 mLab East Africa....................................................................................................................... 73 mLab Southern Africa.............................................................................................................. 83 mLab ECA................................................................................................................................... 93 101 mLab East Asia........................................................................................................................ mLab South Asia..................................................................................................................... 107 Mobile Nepal (mHub Nepal).................................................................................................. 111 TOPICA Mobile Social Networking (mHub Vietnam)....................................................... 117 Akirachix (mHub Kenya)........................................................................................................ 121 Mobile Monday Kampala (mHub Uganda)......................................................................... 126 Appendixes.................................................................................................................................... 134 References.................................................................................................................................... 137 2 7 Lists of Boxes, Figures, and Tables Box 1: Why infoDev?........................................................................................................................... 17 Box 2: The Development Potential of App Economies and infoDev’s Mobile 18 Innovation Program......................................................................................................................... Box 3: What Makes mLabs and mHubs Different from Other Entrepreneurship Support Channels?.......................................................................................................................... 21 34 Box 4: The Role of Innovation Competitions..................................................................................... Box 5: iHub’s Influence, from mLab East Africa’s Perspective........................................................74 Box 6: Noted Startups and App Products Supported by mLab East Africa..................................... 78 Box 7: Noted startups and app products supported by mLab Southern Africa.............................. 85 Box 8: Noted App Products Supported by mLab ECA...................................................................... 96 105 Box 9: Noted Startups and App Products Supported by mLab East Asia...................................... Box 10: Noted Startups and App Products Supported by Mobile Nepal (mHub Nepal)................ 115 Figure 1: Foundational Concepts for mLabs and mHubs................................................................ 19 Figure 2: Mobile Entrepreneurship Enablers at the Center of Mobile Innovation 20 and Entrepreneurship Ecosystems................................................................................................. Figure 3: Location of mLabs............................................................................................................. 23 Figure 4: Location of mHubs............................................................................................................ 23 Figure 5: Comparison of Key Results across mLabs....................................................................... 29 Figure 6: mLabs’ Positioning between an Ecosystem Builder and 33 Startup Accelerator Model.............................................................................................................. Figure 7: Simplified Framework of Tradeoffs for an mLab’s Potential Income Streams................. 41 Figure 8: Environmental Factors and mLabs’ Strategic Positioning as Ecosystem Builders or Startup Accelerators.................................................................................................... 55 Figure 9: mLab Services as Originally Envisaged............................................................................ 59 Figure 10: The Business Model Canvas by Alexander Osterwalder................................................ 71 Figure 11: Business Model Canvas for mLab East Africa................................................................72 Figure 12: Comparative Results for mLab East Africa.....................................................................76 82 Figure 13: Business Model Canvas for mLab Southern Africa........................................................ Figure 14: Comparative Results for mLab Southern Africa.............................................................87 92 Figure 15: Business Model Canvas for mLab ECA........................................................................... Figure 16: Comparative Results for mLab ECA................................................................................96 100 Figure 17: Business Model Canvas for mLab East Asia................................................................. Figure 18: Comparative Results for mLab East Asia..................................................................... 103 110 Figure 19: Business Model Canvas for Mobile Nepal (mHub Nepal)............................................. Figure 20: Business Model Canvas for TOPICA Mobile Social Networking (mHub Vietnam)....... 116 Figure 21: Business Model Canvas for Akirachix (mHub Kenya)................................................... 120 2 8 Table 1: Service Portfolio Options for mLabs and mHubs............................................................... 21 Table 2: Key Funding Amounts and Dates on mLabs and mHubs under CSBKE........................... 22 Table 3: Basic Data for mLab East Africa......................................................................................... 73 76 Table 4: Results Highlights for mLab East Africa............................................................................. Table 5: Budget and Financial Projections for mLab East Africa.................................................... 80 Table 6: Basic Data for mLab Southern Africa................................................................................. 83 Table 7: Results Highlights for mLab Southern Africa.................................................................... 86 Table 8: Budget and Financial Projections for mLab Southern Africa............................................ 90 Table 9: Basic Data for mLab ECA.................................................................................................... 93 Table 10: Results Highlights for mLab ECA..................................................................................... 95 Table 11: Budget and Financial Projections for mLab ECA............................................................. 98 Table 12: Basic data for mLab East Asia........................................................................................ 101 Table 13: Results Highlights for mLab East Asia........................................................................... 103 Table 14: Budget and Financial Projections for mLab East Asia................................................... 106 Table 15: Basic Data for Mobile Nepal (mHub Nepal)................................................................... 111 Table 16: Results Highlights for Mobile Nepal (mHub Nepal)....................................................... 113 115 Table 17: Budget and Financial Projections for Mobile Nepal (mHub Nepal)............................... Table 18: Basic Data for TOPICA MSN (mHub Vietnam)................................................................ 117 118 Table 19: Results highlights for TOPICA MSN (mHub Vietnam)..................................................... Table 20: Basic Data for Akirachix (mHub Kenya).......................................................................... 121 Table 21: Results for Akirachix (mHub Kenya)............................................................................... 123 Table 22: Basic Data for Mobile Monday (MoMo) Kampala (mHub Uganda)................................. 126 Table A.1: Complete mLab and mHub Results data...................................................................... 132 133 Table B.1: List of mLab and mHub Affiliated Startups................................................................... Table C.1: Composition and Definitions of Indicators.................................................................... 136 All dollar amounts are U.S. dollars unless otherwise indicated. 2 9 Executive Summary Part I: Rationale for a Business Model Evaluation of mLabs and mHubs I n 2010, infoDev, the government of Finland, and Nokia launched the Creating Sustainable Businesses in the Knowledge Economy (CSBKE) program to derive and test new approaches to advancing innovation and entrepreneurship in developing countries. Based on the vast growth of mobile content and access technology, a large part of the program focused on support for innovation in mobile applications and software (or “apps”).1 Through CSBKE, the concept of “mobile entrepreneurship enablers” was devel- oped, of which two forms were tested: Mobile Application Laboratories (mLabs) and Mobile Social Networking Hubs (mHubs). mLabs are specialized mobile app business incubation and acceleration service providers; mHubs focus on mobile tech community building by convening stakeholder groups at informal gatherings. mLabs were intended to reach a region comprising several countries, whereas mHubs serve only a single city. Underpinning the mLab and mHub models was the conviction that the enhancement of local innovation ecosystems—characterized by effective partnerships and coordination amongst stakeholders—would improve the enabling environment for mobile entrepreneurship. infoDev also ran a number of global projects (such as innovation competitions) and training sessions, com- plemented by analytical products, such as qualitative evaluations, a major report on user and app ecosystems in developing countries, and a generic business model for mLabs. The total funding for Mobile Innovation programming under CSBKE was about $4.2 million, with about $2.4 million used for grants to local mLab and mHub implementers, $1.5 million for technical assistance for mLabs and mHubs, and $254,000 for research. Four mLabs and eight mHubs were implemented in eleven countries across East Asia, Sub-Saharan Africa, Eastern Europe, Central Asia, and the Caucasus (ECA). An additional mLab was intended for South Asia, but it never became operational because of governance and leadership issues. As of November 2013, periods of operation ranged from over two and half years for mLab East Africa to just over one year for mLab East Asia. As an important step in learning from the mLab and mHub pilot projects, this report analyzes the business models that mLabs and mHubs have implemented, that is, how they generated value for stakeholders and income for themselves. The study derives its findings mainly from in-depth case studies of mLabs and mHubs that were based on almost 150 interviews and 13 focus groups with 240 stakeholders 2 10 held between April and July 2013. The findings will be useful primarily for current and future managers, consortium organizations, and funders of mLabs and mHubs. The report is also relevant for other stakeholders partnering with infoDev in the implementation of mLabs and mHubs, namely private-sector technology partners, investors, and World Bank units, as well as for external practitioners and research- ers interested in technology innovation hubs. The report also lists data for early results of mLabs and mHubs to illustrate their promise. The aggregate data for all mLabs and mHubs as of June 2013 show: • About 2,500 individuals were trained in mobile technology and entrepreneur- ship topics. Over 500 apps were brought to market; more than 200 generated revenue. • million app downloads and more than 5,700 organizational app customers 2.6 were reported. • Nearly 100 startups were created. • About $2.6 million in investments has been raised by the supported startups; together they generated over $1.1 million in revenue and created about 280 direct jobs. The evaluation is complementary to additional analytical products on mLabs and mHubs examining their holistic effects beyond descriptive results, as well as en- trepreneurial success stories. In particular, an outcome assessment of mLab ECA, mLab East Africa, and mLab Southern Africa—to be released in March 2014—finds a significant positive impact of mLabs on the mobile innovation and entrepreneurship ecosystems in which they exist, including improved linkages between ecosystem stakeholders and an overall enhancement of the enabling environment for startup creation. In addition, the outcome assessment documents the development impact of mobile apps developed by mLab-supported entrepreneurs in areas such as ed- ucation, healthcare, financial inclusion, environment, transport, and governance. Part II: Lessons Learned, Discussion, and Future Directions of mLab and mHub Business Models The second part of the report includes the lessons learned and important open questions about the business models of mLabs and mHubs. The lessons learned, as the core contribution of the report, are grouped into four areas: value proposition and services; strategic decisions; financial sustainability; and governance, leader- ship, and consortia. mLabs’ and mHubs’ value proposition and services: • mLabs and mHubs tend to focus on three main areas of activity: startup creation and support, skills development, and community building. • Although mLabs have elements of startup accelerators, they have assumed a broader function as ecosystem builders, focusing on activities that contrib- ute to the enabling environment for mobile entrepreneurs at large, which 2 11 sometimes results in more limited direct effects on startup creation. • Startups see one-on-one mentoring as the key value proposition of mLabs, and they request that it be given stronger emphasis. An mLab or mHub generates the most value when it offers a portfolio of • services that mutually reinforce each other. • Innovation competitions lead to better results for startup creation if they are embedded in continuous support, including follow-up mentoring and contact brokering with potential partners and investors. • Training programs are likely to have positive effects on the local ecosystem in the long run, but they do not appear as the most efficient means for direct startup creation. • mLabs and mHubs add substantial value for entrepreneurs and startups through brokering relationships with clients, funders, and partners. mLabs’ and mHubs’ strategic decisions: • Prioritizing value-maximizing activities has been a major challenge for mLab and mHub managers as they have faced heavy resource and time constraints. • It is important for mLab and mHub managers to engage with the developer and startup community, but mLabs need to evaluate the effectiveness of community building for startup creation and mitigate the risk of distraction. • An mLab’s location must be readily accessible, and colocation with other sup- port organizations can provide the most value for developers an entrepreneurs. • mLabs and mHubs benefit from building partnerships with a wide variety of stakeholders that have a partially overlapping interest in entrepreneurship and startup support. • Events and competitions can be useful tools to build partnerships and select incubatees. • mLab managers are aware that selection of incubatees and tracking them after they leave the program is crucial, but they have sometimes fallen short of doing this adequately. • mLabs have struggled to achieve a strong regional footprint beyond their base country and city. mLabs’ and mHubs’ financial sustainability: mLabs • and mHubs have followed different strategies to work towards finan- cial sustainability; activity-specific funding from tech partners and, for mLab Southern Africa and mLab ECA, funding from provincial and national govern- ments have been the main sources of income. • mLabs and mHubs were successful in securing financial contributions from pri- vate-sector partners primarily for events, competitions, and training, while mLab East Africa has also been successful in seeking sponsorship for its core program. • Only in few cases were mLabs able to attract core funding that would cover the costs of overhead, infrastructure, and incubation from the private sector (for mLab East Africa); major contributions instead came from government 2 12 agencies (for mLab Southern Africa and mLab ECA). • Startups and entrepreneurs are usually open to paying for services or sharing success with mLabs, but they expect a high-value service and are uncertain about the appropriate procedure for far-reaching contractual agreements. • At the end of the infoDev grant period in June 2013, several mHubs and most mLabs, despite progress in attracting revenue from services to private-sector partners and entrepreneurs, project financing gaps over the coming months and years. • Based on the experience of the past two years, infoDev estimates that, with mLabs’ current focus on ecosystem building and support for idea stage enterprises, initial donor financing should extend at least over six to ten years. mLab and mHub governance, leadership, and consortia: • Identifying the right mLab or mHub manager is critical for success. • An mLab consortium should be led by a purpose-driven organization that has an inherent interest in supporting the innovation ecosystem. • Consortia that are led by government-linked entities or universities often suffer from bureaucratic requirements that can conflict with the mission and mode of operation required for an effective mLab. • Consortia leaders need to have sufficient institutional capacity and set clear expectations and roles for consortia partners. • Each consortium organization needs to ensure internal stability and long- term backing of the mLab and its vision; risks of internal challenges and bottlenecks have to be mitigated early on. • Consortium member organizations are important channels for resources and partnerships for an mLab; the organizations ought to complement each other in this respect. • If consortium organizations directly contribute to the implementation of the mLab’s activities, the consortium needs to be able to manage potential 2 13 conflicts of interest. Beyond findings, the report also highlights several unresolved questions about mLab business models. For example, should mLabs focus more on ecosystem building or startup acceleration? At which innovation and startup stage can mLabs add the most value? Should mLabs be purely market-oriented or also accommo- date social enterprises? Which areas of expertise should mLabs emphasize as their unique value proposition in innovation ecosystems? Should mLabs build a portfolio of incubatees with complementary assets? And should mLabs make direct equity investments or rather broker investor contacts for client startups? The report also discusses lessons learned and future directions for infoDev. In par- ticular, it describes how enhanced scoping assessments and ecosystem mapping will enable the design of appropriate mobile entrepreneurship enabler models in future programs, and how improvements can be made with regard to infoDev’s future programs’ timeline and grant administration. Further suggestions concern infoDev’s increasing role as a knowledge and analytics provider for mLabs and mHubs, as well as improved ways for global network facilitation. Finally, important knowledge gaps are highlighted that could be addressed through future research by infoDev or other organizations working in the area of innovation ecosystems of developing countries. Part II of the report concludes that the mLab and mHub pilots have established proof for the viability of the mobile entrepreneurship enabler concept: mLabs and mHubs are capable of creating and supporting growth-oriented startups and filling gaps in innovation ecosystems of developing countries. At the same time, continuous learning will be needed for mLabs, mHubs, and infoDev to maximize 2 14 their effectiveness. Depending on the local context and strategic choices, mLabs and mHubs can implement a variety of business models, and experimentation and adaptation will remain necessary. In particular mLabs’ revenue and financial sustainability models require further testing and adjustment. The lessons drawn from the CSBKE pilots will not only help infoDev to deepen its continuous value proposition for its key stakeholders, but they will also guide the design of future activities included in the infoDev work program, such as infoDev’s upcoming regional Mobile Innovation Project in the Caribbean or its inclusion-ori- ented activities planned for Africa. Within the broader World Bank context, these findings will also be valuable to several projects that encompass models inspired by the mHubs and mLabs. Part III: Case Studies of mLab and mHub Business Models Case studies, as Part III of the report, provide additional information for stake- holders in the countries and regions of implementation. They should also give useful pointers and ideas for improved business model design of mLabs and mHubs as well as similar technology innovation hubs. The Business Model Can- vas serves as the principal analytical framework. In seven in-depth case studies, the report derives rich illustrations and descriptions of the business models of mLabs and mHubs and considers the challenges they face, the results achieved, as well as their longer-term financial sustainability. The report covers mLab East Africa (Kenya), mLab Southern Africa (South Afri- ca), mLab ECA (Armenia), mLab East Asia (Vietnam), Mobile Nepal (mHub Nepal), TOPICA MSN (mHub Vietnam), and Akirachix (mHub Kenya). Essays on mLab South Asia (Pakistan) and Mobile Monday Kampala (mHub Uganda), which were based on interviews only with key implementers and decision makers, are included in addi- tion to the in-depth studies, for a total of nine case studies. For reference, mHubs in Azerbaijan, Georgia, Moldova, and Tanzania were not covered as case studies. Endnotes Another key activity pillar of the CSBKE program focused on agribusiness. See http://www.infodev.org/workprogram 1 2 15 for an overview. Part I: Rationale for a Business Model Evaluation of mLabs and mHubs Introduction and Background and innovation networks can be the D necessary condition for a long-term eveloping countries face tre- upward spiral of learning and innovation mendous structural barriers to processes to emerge (Feldman, Francis, economic growth, and gaps in and Bercovitz 2005; Martin and Sunley skills, financing, and infrastructure 2003; and Meyer-Stamer 2002). put a substantial strain on productive, competitive business activity. This is The CSBKE Program even more so the case in local knowl- In 2010, infoDev, the government of edge economies, where output only Finland, and Nokia took aim at tackling marginally depends on material inputs structural barriers to technology inno- but rather on people’s ingenuity and vation in developing countries through networks. Much has been written about the Creating Sustainable Businesses the high and rising contribution of the in the Knowledge Economy (CSBKE) knowledge economy to total economic Program. The program is informed by a activity and growth, and it has been philosophy that acknowledges the com- established that technology innovation plex interactions in local knowledge and growth-oriented entrepreneurship economies, making use of the notion of are driving forces in this context (Wong, innovation ecosystems. Ho, and Autio 2005).2 infoDev sought to leverage its institu- Importantly, gaps but also contributors tional knowledge on business incubation to technology innovation are often inter- and information and communication dependent and perpetuate themselves. technology (ICT) to strengthen the mo- As a simple example: if there are no bile innovation ecosystems of developing innovators that start successful busi- countries. Moreover, infoDev seemed nesses, or if incumbents are not willing well-situated to implement CSBKE to mentor their peers, professional con- because of its unique positioning and tacts and angel investments will be more mission: as a trust-funded program limited for younger innovators, which within the World Bank, infoDev has the in turn limits their potential to become mandate to pilot and experiment inno- role models for following generations vative programs, and it can engage di- of entrepreneurs. Often, the existence rectly with private sector and grassroots 2 16 of at least rudimentary entrepreneurial organizations in developing countries, but also access the World Bank’s infra- ecosystems of partner countries that structure, networks, and brand (see box would benefit the broader economy 1). The program was also in line with and society, job creation through small Finland’s development policy, at the time and medium enterprises (SMEs), as focusing on private sector development well as improved knowledge on the and fostering information societies.3 opportunities of business incubation and mobile apps to enhance the pro- The overarching ambition of CSBKE ductivity of SMEs. To achieve such was to bring about positive changes in far-reaching goals, it became clear the innovation and entrepreneurship that the funding should not simply Why infoDev? “infoDev is privileged to work with a range of donors and partners who value our grassroots innovation communities and our cutting-edge approach to piloting new ideas and seeking new knowledge. […] infoDev has the necessary agility to pilot new concepts at the grassroots, and to ensure that workable solutions are scaled and mainstreamed into larger projects. It supports the World Bank Group’s commitment to innovation and entrepreneur- ship through these approaches. It also bridges operations and knowledge through best-practice assessments, research, and publications, and through a regular and rigorous evaluation of its activities. infoDev supports the World Bank Group’s commitment to the growth of a strong private sector in developing countries in a manner that leverages technology and innovation and feeds overall growth, competitiveness and inclusion. Our work extends around the world, allowing us to leverage the ideas and lessons generated in one country for the benefit of other countries. Our stakeholders in developing countries create and design practical solutions with infoDev and work with us as partners in ensuring success on the ground. These stakeholders and our networks also request reliable and rigorous research and participate in infoDev’s research and knowledge agenda. This research, which draws upon grassroots voices, in turn informs effective operations. […] Over the past few years, infoDev has built on its success in incubating technol- ogy-enabled businesses to launch specialized programs aimed at promoting the growth of new ventures in the mobile, climate and agribusiness sectors. Supporting these strategic sectors contributes to broader growth and compet- itiveness, and leads to the development of value-adding jobs suited to the new knowledge economy.” Source: infoDev Work Program 2013-15, www.infodev.org/workprogram Box 1: Why 2 17 infoDev? be infused in local economies of the Branscomb 2003; Baird, Bowles, and nineteen countries of implementation. Lall 2013; and Heydebreck, Klofsten, Rather, projects were intended to stimu- and Maier 2000). Second, the program late and catalyze locally driven, sustained did not emphasize subsistence and dynamics in innovation ecosystems. necessity entrepreneurship (including micro-entrepreneurship), but instead The program advocated piloting of high targeted growth-oriented entrepre- potential development interventions, neurs as potentially the most signifi- coupled with continuous evaluation and cant contributors to overall economic learning, as the best way to push the growth and transformation (Autio 2008; emerging field of innovation ecosystem and Fritsch 2008). support. First, this led to a focus on the early stages of innovation, where, argu- ably, the most potential for technology A Public-Private Partnership Program innovation is yet untapped because for Mobile Entrepreneurship Enablers few existing support institutions cover infoDev and the government of Finland this “pioneer gap” (Auerswald and opted to focus a substantial share of The Development Potential of App Economies and infoDev’s Mobile Innovation Program I nnovation and entrepreneurship are key drivers of development. Among the most vibrant areas for entrepreneurship are mobile applications and software, or simply “apps.” The growth of “app economies” is forecast to further increase over the coming years, implying enormous potential for app entrepreneurs and developers. Mobile apps also enhance opportunities in areas such as healthcare, government, empowerment, and entertainment. App entrepreneurs face substantial challenges in fledgling innovation ecosys- tems of developing countries. They might have unique insights into local cir- cumstances and abundant creativity but they often lack professional networks, favorable policy environments, information access, business skills, mentors, spaces for exchange, and access to investors. In most cases, ecosystems also have to integrate policy makers, mobile network operators, investors, donors, and so on, for mobile app entrepreneurs to thrive. infoDev’s Mobile Innovation Program supports growth-oriented mobile app businesses in emerging and frontier markets. infoDev does this by 1) enabling Box 2: The entrepreneurship in the mobile industry, through venture incubation and accel- Development eration, 2) building mobile innovation communities of entrepreneurs, investors, Potential of App partners, and mentors, 3) researching the app economy of local markets. Economies and infoDev’s Mobile For more information on “mobile innovation ecosystems” and “app economies,” see www. Innovation worldbank.org/ict/ic4d2012, http://unctad.org/en/PublicationsLibrary/ier2012_en.pdf, and 18 Program http://www.visionmobile.com/product/app-economy-forecasts-2013-2016/ Innovation ecosystems Mobile app Tech & startup economies communities Figure 1: Business mLabs Startup Foundational incubation acceleration and mHubs Concepts for mLabs and mHubs CSBKE programming on the mobile recruitment decisions of managers and app4 industry, which later led to the provision or facilitation of core funding. establishment of the Mobile Innovation infoDev and Nokia agreed that any tech Program as one of infoDev’s work partnership would be non-exclusive to- streams (see box 2). The program was wards other mobile platform providers. set up in a public-private partnership This focus on the mobile app industry with Nokia—one of the leading mobile was made to tap into the tremendous communications equipment suppliers potential of growth in app economies to for developing countries. contribute to overall economic growth and development, which had begun to Participation in the program was in become more apparent towards the line with Nokia’s goals as it hoped to end of the last decade. Moreover, a fo- foster innovation in mobile content cus on mobile apps also built on each ecosystems and support app develop- partner’s strengths and on the large im- ers generally, while ensuring that its pact potential for innovation ecosystem technology would play an important support in the mobile sector (low capital role in the process. Nokia’s main role requirements and barriers to entry, a was seen in bringing in technical and fast growing distribution platform of market expertise and strengthening the mobile infrastructure and data-enabled program’s private-sector orientation. devices, as well as an emerging skills The company played a strong role espe- base in app development). cially in the setup phases: it helped con- ceive the concept and general business The backbone of the Mobile Innovation model for mLabs (see below for details), Program would be built by implemen- and the company also contributed to the tation partners on the ground. This was consortium selections. Nokia’s partic- seen as the best way to enable contex- ipation was also expected to enhance tualization and adaptation to the evolv- the value proposition of projects mainly ing conditions of innovation ecosys- through in-kind support from its local tems, and to contribute to sustainable subsidiaries, for instance, in the form of development of local knowledge econ- training material, software toolkits for omies. In the design of potential im- app development, knowledge sharing plementation models, infoDev took the with project managers and clients, as familiar concept of business incubation well as support for program scoping and integrated it with design elements and design. At some of the sites, Nokia drawn from areas such as startup ac- would also be involved in the later stag- celeration5, app economies6, and tech es of project rollout, for example, in and startup communities7(see figure 1). 19 B I L E M O • I N M Mobile N technology O E companies V T A ICT E C O S Y S Early stage infrastructure T I investors Mobile O N entrepreneurship enablers Entrepreneurship Policy & business A N support (incubators, mentors, coaches, serial entrepeneurs) environment Figure 2: Mobile P Talent, skill, D Entrepreneurship I and mindset H Enablers at the (secondary & E tertiary education, S Center of Mobile training) N R Innovation and U R T Entrepreneurship E P R E N E Ecosystems This implied that infoDev would move InfoDev aimed to establish mLabs from traditional incubators towards and mHubs through a market-driven more flexible mobile entrepreneur- approach, achieved via the global ship enablers that would create and partnership with Nokia and a focus support startup companies, but also mainly on nongovernmental local help to grow the entire local mobile implementation partners. This was innovation and entrepreneurship eco- seen as the basis to push and support system (see figure2).8 entrepreneurs to generate viable, sus- tainable mobile app business models, What Are mLabs and mHubs? fit for the global marketplace or highly infoDev decided to experiment with two relevant locally. mLabs and mHubs different models of mobile entrepre- were expected to become important neurship enablers, mobile application contributors to the growth of com- labs, or mLabs, and mobile social net- petitive mobile app industries. Their working hubs, or mHubs (see table 1 for market orientation was envisioned as a list of services). the foundation for mLabs and mHubs mLabs are specialized mobile business to attract strong participation and en- incubation and acceleration facilities, dorsement by private-sector technol- offering physical workspaces, mentor- ogy companies and other stakeholders ing and coaching, devices for app-test- of local innovation ecosystems. ing, training, and startup competitions. Typically, mLabs would only be imple- mHubs build mobile tech communities mented if the local innovation ecosystem by convening a variety of stakehold- had already reached a certain degree er groups at informal gatherings, of maturity. When critical inputs such peer-learning sessions, confer- as entrepreneurial and technical tal- ences, and ideation and prototyping ent, early-stage investors, and strong 20 competitions. partners with support resources are What makes mLabs and mHubs different from other entrepreneurship support channels? • Focus on growth entrepreneurship and startups • Focus on venture creation and early stage innovation support (“From Mind to Market”) • Expertise in mobile tech and apps as attractive growth sectors with particular market dynamics • Rooted in notion of complex multistakeholder ecosystems (innovation ecosystems, entrepreneurship ecosystems, startup ecosystems, mobile tech ecosystems) • Participatory, demand-driven codesign by infoDev and local stakeholders • Independent implementation through grantees as local mobile entrepreneurship enablers; leeway to fill out a conceptual framework provided by infoDev Box 3: What Makes mLabs • Local enabler has access to global infoDev network (peer enablers, knowledge, and mHubs experts, and so on) Different from • Local enabler has indirect access to World Bank infrastructure, branding, Other Entrepre- and networks neurship Support Channels? unavailable, mLabs can hardly real- with small amounts of resources and ize their full potential. mLabs have a funding, which means that they can fit regional focus, beyond their host city into even the most nascent innovation and country. While mHubs also rely ecosystems. In sum, mLabs and mHubs on partner organizations as sponsors have several attributes that differentiate and in-kind contributors, they can them from other entrepreneurship and form communities from the ground up business support organizations (see box 3). Table 1: Service Portfolio Options mHubs mLabs Service Portfolio Informal community building and networking events Options for (e.g., thematic presentations & dicussions, showcasing events) mLabs and Facilitation of online collaboration and learning mHubs Multi-stakeholder conferences Mentorship by successful entrepreneurs and investors* Technical and business-skill training* Innovation competitions and prototyping events like hackathons for talent sourcing* Assistance with marketing apps, e.g., advice on pricing, facilitating delivery channels* *Advanced mHubs can Contacts to investor networks* implement this activity, for instance, assisted Provision of co-creation and ideation spaces by an mLab. Formalized incubation program **mLabs can only Free/subsidized office spaces provide direct access Mobile app testing facilities to seed capital if they receive additional Access to early stage financing and seed-funding** 21 funding. Funding Total CSBKE $19,500,000 Mobile Innovation programming under CSBKE $4,171,000 Funding for mLabs and mHubs $3,917,000 Grants $2,439,000 Technical assistance $1,478,000 Research a $254,000 Dates CSBKE launch December 2009 Table 2:: First tranche received by infoDev/the World Bank April 2010 Key Funding First selection process (for mLabs in Africa) February to September 2010 Amounts First activities launched by an mLab or mHub (mHub Uganda) February 2011 and Dates on First official launch of an mLab (mLab East Africa) June 2011 mLabs and mHubs under Most recent official launch of an mLab (mLab East Asia) September 2012 CSBKE End of mLab grant periods June 2013 Note: Funding amounts are based on expenses as of September 10, 2013, excluding few outstanding commitments. Amounts have been rounded to thousands. a. CSBKE research outputs that fall under the Mobile Innovation Program include the IC4D2012 – Maximizing Mobile report as well as an overview report and country studies on South Africa and Kenya on the topic of Mobile Usage at the Base of the Pyramid. Funding and Rollout of mLabs (ECA). Shortly after the trust fund had and mHubs under CSBKE been established, infoDev, the govern- ment of Finland, and Nokia reached About $2.4 million of CSBKE fund- consensus that a wider spread of the ing was used for recipient-executed network—which could enhance its activities in the form of grants to reach and the potential for experi- local partners that would implement mentation—would be feasible with mLabs and mHubs. An additional $1.5 the available funding. Accordingly, five million was spent on technical assis- mLabs would now be implemented, tance for mLabs and mHubs, executed one in each the following regions: by infoDev. The overall duration of East Africa, Southern Africa, ECA, CSBKE was four and half years, from East Asia, and South Asia (see figure December 2009 to June 2014. Grant 3). However, the mLab intended for periods, which imply the grantees’ South Asia never became operational timeframe for implementation, were because of governance and leadership scheduled to last from mid-2011 to issues (see essay on mLab South mid-2013, while mLab East Asia was Asia). With regard to mHubs, three officially launched only in September each were funded for the Africa and 2012 (see table 2). ECA regions, as well as two in Asia, for a total of eight (see figure 4). infoDev currently has four mLabs in countries in the developing world.9 Given their pilot character, it was diffi- Originally, three mLabs were due to be cult to set concrete targets for mLabs launched, one each in Africa; Asia; and and mHubs, both at the program and 22 Europe, Caucasus, and Central Asia project level. Originally, each mLab Global Partner: Moldovian Technology Georgian Business Caspian Investment Young Innovations, Mobile Monday Transfer Network, Development Council, Center, Baku, Kathmandu, Nepal Chisinau, Georgia Tbilisi, Georgia Azerbaijan • •• • MoMo CRC Topiaca, Kampala, Uganda • Hanoi, Vietnam •• • Akirachix COSTECH, Nairobi, Kenya Dar-es-Salaam, Tanzania E I G H T m H U B S S E L E CT E D FROM 40 APPLICANTS Figure 3: Location of mHubs mLab South Asia iHub consortium EIF (Armenia) PSEB consortium (Kenya) (Pakistan) • • • • • Meraka consortium SHTP consortium (South Africa) (Vietnam) FIRST FIVE mLABS S E L E CT E D F R O M 7 5 A P P L I C A N T S 23 Figure 4: Location of mLabs was expected to incubate ten mobile pilot character and catalyst function application entrepreneurs. The CSBKE of mLabs and mHubs had always work program also specified, in hind- informed the goals of the program, sight, a quite modest outcome target namely to experiment with viable for the development and success of business models for mobile entre- apps through mLabs and mHubs: eight preneurship enablers and explore the to ten mobile apps with potential to overarching impact that they could enhance productivity in local knowl- have on local innovation ecosystems edge economies were expected to be and knowledge economies. launched and scaled by each mLab. Later, grant agreements with mLabs Similarly, at the outset of the CSBKE, it specified additional targets, mostly was unclear if mLabs and mHubs would at the output or immediate outcome be able to reach financial sustainability level, such as the number of students within the program timeframe of about that should be trained or the number of three years, with just over two years of teams that should be incubated. Most implementation. While program de- mHubs were not given explicit targets signers hoped that mLabs and mHubs other than the number of events they would be able to attract substantial were expected to organize. financing from private-sector entities, follow-on funding from other donors Such targets had the function to en- and local governments had always sure accountability for deliverables but been considered as a likely option. they were not seen as comprehensive Ultimately, infoDev, the government of indicators for the overall impact that Finland, and Nokia—although aware 24 mLabs and mHubs would have. The that not all mLabs and mHubs might achieve sustainability by the end of the CSBKE program—decided that it was paramount to get projects off the ground as quickly as possible. The hope was that additional funding would be generated once any local business models’ promise would start to materi- alize and proof of concept (locally and/ or globally) would be reached. Despite mLabs’ and mHubs’ opera- tional independence, they cannot be seen in isolation from infoDev as their global umbrella organization. At the beginning of implementation, infoDev provided in-depth conceptual guid- ance to new mLab managers by com- missioning a generic mLab business plan10, presentation material on the mLab concept11, and a business plan • Facilitating partnership building, workbook. These resources made the especially with global technology case for mobile technology-specific and development partners.13 incubators, outlined potential service • Evaluating mLab and mHub pilots portfolios and partnership models, and to generate business analytics and gave ideas for revenue generation. At good implementation practices. the same time, it was emphasized that • Enabling and encouraging peer mLabs would have to adapt this gener- learning and knowledge exchange ic model to local circumstances. between mLabs and mHubs. Over the course of the rollout of CSBKE, infoDev had an interest in Why Evaluate the Business Models seeking information from the local lev- of mLabs and mHubs? el in order to report and showcase the The pilot character of mLabs and results of its programming and to learn mHubs implies the appropriate mea- for future projects. More importantly, sures for success. Of course, any given infoDev also aimed to create a network mLab or mHub has the goal to increase of mobile entrepreneurship enablers its clients’ business success, skills, that was more valuable than the sum and professional networks, and this of its parts. To do this, infoDev explored can and should be tracked in numbers. the following: While quantitative results cannot tell • Running global activities such as the whole story of technology innova- innovation competitions and hack- tion hubs such as mLabs and mHubs, athons, with mLabs and mHubs as quantitative measures are relevant and local implementation partners.12 25 will increase in importance as more sophisticated benchmarks and assess- in which they exist. None of the mLabs ment methods become available over and mHubs have been analyzed by the coming years. infoDev encourages means of rigorous impact evaluation mLabs to systematically track and methods that could more clearly attri- communicate the success of their client bute impact and quantify any effects. startups over time as such transparen- Hence, it is misleading to assume that cy is critical for incubators and accel- there is a linear relation between stan- erators to continuously attract interest dardized observed indicators and per- from new startups and investors.14 formance for a given mLab or mHub. It should not be inferred that results are a This report does not claim to compre- direct consequence of only the mLabs’ hensively assess the holistic impact of and mHubs’ activity, so, for instance, mLabs and mHubs, but rather, it aims the results cannot lead to a clear con- to identify models and lessons that ap- clusion such as “mLab East Africa was pear likely to lead towards the largest the most successful because it had the possible catalytic impact.15 It makes best results.” this assessment based on infoDev’s experience and the qualitative evidence Second, in the programmatic context of collected from stakeholder interviews, CSBKE, mLabs and mHubs were pilot reflected against the current under- projects. The fundamental purpose of standing of startup ecosystems and the the mLab and mHub projects was to role of (grassroots and tech) innovation get them to a proof of concept stage, for economic growth. upon which working models could be scaled or replicated elsewhere. The In order to do this, the report analyzes underlying goal was to push frontiers mLabs’ and mHubs’ business models. in innovation for development practice “Business model” is defined broadly, and to identify mobile entrepreneurship and encompasses how mLabs and mHubs “make money” but also how enabler models that work in complex they generate value for various stake- multistakeholder ecosystems. While holder groups, even if this does not direct positive effects on private sector translate into an immediate revenue development were a key goal, concrete stream. infoDev collected initial quan- impact targets were deliberately held titative data, but mLabs and mHubs to a modest level. mLabs and mHubs should not be evaluated uniquely based were allowed to try out different mod- on such information on startup revenue els that could lead to the highest ho- generation, startup creation, individ- listic impact, without them being held uals trained, and so forth. A strong accountable to extensive impact goals focus on models and a lesser focus on such as startup revenue or job creation. quantitative data are mandated by the Third and most importantly, mLabs and following reasons. mHubs are catalytic, innovation eco- First, mLabs’ and mHubs’ outcomes system-oriented private sector devel- are highly interdependent with the dy- opment instruments, and their power 26 namics of the innovation ecosystems is not expected to lie in the immediate also create business connections that otherwise would not have been created, and there are many more examples of An mLab’s or mHub’s indirect positive effects. holistic impact on the local innovation and Accordingly, this study will list quanti- startup ecosystem tative results of mLabs and mHubs, but only insofar as they help to understand goes much further the viability of their business models. than what short-term Also, this analysis does not aim to yield results can capture. a definite, universal mLab or mHub Entrepreneurs learn business model. Local conditions and constraints will always dictate, for and get inspired, and example, which services add the most even if they do not start value or which types of partner organi- a business right away zations can meaningfully contribute. or if their business Instead, the study carves out high-level fails, they might found lessons across contexts, and highlights a successful company which choices mLab and mHub manag- at a later point in time. ers can make to adjust their business model to their local context or to fulfill a specific mission focus. Case studies further complement high-level findings results they produce, even if they reach through illustrations of context-specif- the proof of concept stage. The support ic implementation opportunities and of ecosystems and communities cannot challenges. Through this combination, be understood as a mechanistic relation the report provides suggestions and between input and output. An mLab’s or ideas for business model design and mHub’s holistic impact on the local in- conceptualization of innovation hubs in novation and startup ecosystem is likely general and new and existing mLabs to go much further than what descrip- and mHubs in particular. The findings tive, short-term results can capture.16 were derived from 148 interviews and Coupling resources with a culture that 13 focus groups with 240 stakeholders embraces tech innovation and a startup of these seven mLabs and mHubs. mentality can ultimately create fruitful “co-opetitive” exchanges and positive This part of the report (part I) concludes feedback cycles in ecosystems, as ex- by describing aggregate results data amples such as Boulder,17 Bangalore,18 that illustrates the promise of mLabs and Tel Aviv19 have shown. For instance, and mHubs. A more substantive anal- entrepreneurs learn and get inspired, ysis is included in part II of the report, and even if they do not start a business which summarizes generalized les- right away or if their business fails, they sons learned as well as discussions might found a successful company at a of inconclusive evidence and future later point in time. mLabs and mHubs 27 directions for infoDev. In part III, the report employs rich illustrations and • More than 280 direct high-quality descriptions of nine case studies (out of jobs have been created.20 a total twelve mLab and mHub pilots), • Almost 100 startups have been including seven in-depth discussions of created and close to 300 startup mLabs’ and mHubs’ business models. teams have received mentoring and coaching. The report’s findings will be useful for • Over 500 apps have been brought the following audiences: to market, of which more than 300 • Current and future mLab and have made revenue. • 2.6 million app downloads21 and mHub managers and consortium organizations. more than 5,600 organizational app • Current and future stakeholders customers22 were reported. • About 370 events have been orga- and partners of mLabs and mHubs. • Current and future private-sector nized (such as startup events and partners of mHubs, mLabs, and app contests) and 2,500 individuals infoDev (in particular, mobile tech have been trained in mobile tech- companies). nology and entrepreneurship topics. • Overall, an estimated 32,000 • Donors and funders for mLabs and mHubs (including local government developers and entrepreneurs agencies). have been reached. • Other entrepreneurship enablers Such quantitative evidence, if contextu- (such as innovation hubs, incuba- alized and compared properly, can be tors, and accelerators). useful to understand the potential of • Donors and funders of infoDev’s mLabs and mHubs, to learn about and Mobile Innovation Program. improve their operations, and to show- • Researchers and practitioners case and promote them to funders working in the innovation for and partners. When comparing mLabs development space. against each other, it clearly emerges • infoDev and other World Bank units that, to date, mLab East Africa has working on tech innovation support. produced the most promising results, with the highest score in nine out of the Results for mLabs and mHubs eleven indicators analyzed (see figure 5). This certainly reflects how active infoDev has collected data for mLabs this mLab has been. mLab East Africa and mHubs directly from their manag- has pursued an expansive strategy, im- ers. Key results and outputs reported plementing virtually every service line by the four operational mLabs and eight possible for an mLab, while maintain- mHubs are summarized below: ing a high pace (see case study on mLab • About $2.6 million has been raised East Africa). At the same time, without by startups supported by mLabs taking credit away from mLab East and mHubs. Africa’s achievements, these scores • The startups together have gener- are only indicative measures for suc- 28 ated over $1.1 million in revenue. cess (see appendix C for details). For Overall Revenue generated reach by startups 100% 80% 60% Investments Events raised by startups 40% 20% 0% Applications Jobs created mLab East Africa for support & supported mLab Southern Africa mlab ECA mlab East Asia Training Startups & startup teams created In-depth App customer team support traction App innovation, (apps monetized, brought to Figure 5: market, prototyped) Comparison of Key Results Note: The percentage values represent linear comparisons across data for all four mLabs, with the mLab scoring across mLabs highest for a given indicator receiving 100%. Appendix C provides details on the calculation methods example, mLab East Africa has been strategy shift towards a focused, verti- operational for the longest time and it cal mobile incubator model (see case has not faced implementation delays study on mLab Southern Africa) has or management changes as mLab resulted in a relatively low number of East Asia and mLab Southern Africa people trained and teams supported, have. mLab East Africa also thrives on while those startups that have been the buzz of the Nairobi ecosystem and supported were successful, generat- high demand for its services. ing fairly high numbers for startup rev- enues and investments raised. Finally, Interestingly, some of the strategic mLab East Asia—after a somewhat priorities of the other three mLabs bumpy path to implementation—has are reflected in the comparative re- reached the largest number of devel- sults. For instance, mLab ECA’s long opers and entrepreneurs and trained term approach to support talented many individuals in a short period (see developers with no entrepreneurial case study on mLab East Asia), but the experience from the ground up has startup teams will need more time and yielded extremely popular consumer in-depth support until they are able to apps, while the results for startup raise substantial investments and gen- revenues and investments raised have erate revenue. remained modest after the first year of implementation (see case study on mLab ECA and At Which Startup Stage Can mLabs Add the Most Value?). On 29 the contrary, mLab Southern Africa’s Endnotes 2 Or see http://www.nytimes.com/2013/08/18/business/why-innovation-is-still-capitalisms-star.html. 3 See, for instance, Finland’s Aid for Trade Action Plan (2008–2011), http://formin.finland.fi/public/default.aspx?contentid=137494&contentlan=2&culture=en-US. 4 In this report, “app” is defined broadly and includes any mobile application, software, or content on any vendor platform, including USSD and SMS-based technologies. 5 Startup acceleration is distinct from business incubation through a greater emphasis on investment-readiness, mentoring and coaching, peer learning within badges of entrepreneurs, as well as pitching and demos. In contrast to incubation, acceleration de-emphasizes physical work and office space, provision of business support functions such as accounting and legal, and other physical infrastructure such as app testing facilities. 6 See, for instance, World Bank, 2012; UNCTAD, 2012; or http://www.visionmobile.com/product/ app-economy-forecasts-2013-2016/. 7 See, for instance, http://steveblank.com/2012/10/08/startup-communities-regional-clusters/. 8 This report defines mobile innovation and entrepreneurship ecosystem broadly, encompassing all factors mentioned in figure 2. 9 mLab regions are not equivalent to World Bank regions and instead cover subregions of neighboring or otherwise geographically/culturally/politically related countries. 10 http://www.infodev.org/infodev-files/resource/InfodevDocuments_1087.pdf. 11 http://www.vitalwaveconsulting.com/pdf/2011/mLabs-Speaker-Series.pdf. 12 infoDev implemented the m2Work project (http://www.infodev.org/m2work), consisting of a global idea competition (https://ideasproject.com/web/m2work) and a multisite hackathon (http://m2workhackathon.org/); the mAgri Challenge (http://www.infodev.org/mAgri), and the VentureOut challenge (http://ventureoutchallenge.org/). 13 For instance, together with Nokia, infoDev facilitated a partnership between AppCampus and the mLabs, according to which mLab startups applying for the program would have privileged access to earmarked funds, see http://infodev.org/ press-releases/nokia-world-bank%E2%80%99s-infodev-and-appcampus-announce-collaboration-press-release. 14 For a recent op-ed on the topic by David Cohen, founder of Techstars, see http://blogs.wsj.com/accelerators/2013/08/20/ david-cohen-entrepreneurs-deserve-full-transparency/. The importance of continuous tracking of client startups for im- pact accelerators was also highlighted in a recent report by the Aspen Network of Development Entrepreneurs (see Baird, Bowles, and Lall, 2013). The study gives the following advice to impact accelerators: “Tracking & Measurability: [1.] Invest in platforms and systems to encourage and enable quality data collection from the enterprises you support. (Check out the free PULSE platform on Salesforce.) [2.] Collect data from all enterprises that apply to your programs, even the ones that are not accepted or do not receive services, to more comprehensively assess performance against a control group. [3.] Collect data from participating enterprises for at least five years post-program years to track progress and growth over the long-term. Partner with academic institutions and industry associations to develop stronger data collection systems.” 15 An upcoming outcome assessment of mLabs and mHubs implemented under CSBKE will evaluate additional qualitative and quantitative evidence. infoDev commissioned the assessment from the global private sector development evaluation and research agency, CAD (http://cad.globalcad.org/en/que-es-cad/). 16 For a strong rebuttal of the usefulness of short-term quantitative indicators as measures of a startup ecosystem’s success, see this interview with Brad Feld http://www.technologyreview.com/news/516521/its-up-to-you-entrepreneurs/. 17 See, for instance, http://www.kauffman.org/newsroom/2013/08/ young-hightech-firms-outpace-private-sector-job-creation. 18 See, for instance, http://www.forbes.com/sites/quora/2012/12/07/how-can-a-young-twenty-something-american-find- meaningful-work-in-indias-tech-startup-scene/ or http://tech.co/india-startup-scene-shradha-sharma-2012-05 19 See, for instance, http://startupnationbook.com/. 20 Some of these jobs are employees in fledgling incubatee startups. More reliable figures for sustainable job creation will only become available in the coming months and years. 21 Downloaded apps only include apps for which downloads were tracked and the actual number is likely to be much higher. 30 22 Organizational customers include enterprises, public agencies, and other organizations. 31 Part II: Lessons Learned and Future Directions for mLab and mHub Business Models Lessons Learned for mLabs (1) Startup creation and support through and mHubs incubation and acceleration, (2) Skills development through training T he mLab and mHub pilots analyzed and workshops, and showed a set of commonalities. (3) Community building through events This section summarizes findings and online platforms. that appear to hold across local contexts, Startup and innovation competitions or that seem to be rooted in general contributed more or less to all three trends and attributes of mobile innova- areas, depending on their design. tion ecosystems in developing countries. Although mLabs have elements of start- These generalized findings should not up accelerators, they have assumed be seen to imply a set of definite best a broader function as ecosystems practices. Of course, each mLab and builders, focusing on activities that con- mHub is designed and implemented tribute to the enabling environment for differently, and responsiveness and mobile entrepreneurs at large, which adaptation to local innovation eco- sometimes results in more limited di- systems and market conditions are rect effects on startup creation. While paramount. Still, if applied to local startup creation and support have contexts with care, the findings and been important goals for mLabs, they lessons should provide useful guid- have also assumed a broader func- ance for mLab and mHub managers, tion as enhancers of local innovation infoDev and donor organizations, as ecosystems. In part, this was a direct well as partners and other ecosystem consequence of the original mobile stakeholders. At the least, they can entrepreneurship enabler concept as help to avoid repetition of mistakes imagined by infoDev. mLabs have de- that were made and set more realistic livered training and community build- expectations. ing, even if there were no direct or immediate effects on startup creation. Services and Value Proposition What is more, mLabs have helped en- mLabs and mHubs focused on three trepreneurs from the early stages on- areas of activity, with varying emphasis wards; often mLabs supported recent depending on the local strategy and university graduates with not much 2 32 ecosystem requirements: more than an idea—enabling them to Ecosystem mHubs mLabs Startup builder accelerator Increase Main Goal Community startup building valuation Secondary Enhance skills Accelerate goal & experience startup growth Minimum innovative Startup Idea stage stage with traction Number & type Many Few of clients eligible individuals incorporated ©©© for support and teams startups Infrastructure Key resources & convening Mentors space Broad (multi- Narrow Approach to stakeholder) (access networking to investors) Figure 6: build startups from the ground up— linkages between ecosystem stake- mLabs’ Positioning even if this implied a higher failure rate holders), and not primarily in startup between an and more modest business success- creation. A simplified framework that Ecosystem Builder es. Accordingly, many interviewees depicts key differences in the goals and Startup considered mLabs’ greatest impact to and support strategies between an Accelerator Model lie in ecosystem building (nurturing ecosystem builder and a startup ac- talent and ideas, teaching business celerator model is presented in figure and technical skills, and enhancing 6. It remains an open question wheth- er mLabs should shift their emphasis towards startup acceleration or main- tain services that enhance the innova- tion ecosystem at large (see section Startups see one- Should mLabs Focus on Ecosystem on-one mentoring Building or Startup Acceleration?). as the key value proposition of mLabs, Startups see one-on-one mentoring as the key value proposition of mLabs, and and they request that they request that it be given stronger it be given stronger emphasis. Early-stage startup entre- emphasis. Yet, it is preneurs seek feedback on common usually hard to find business issues (such as negotiations, partnership building, fundraising, good mentors, even business planning, financial mod- for payment. eling, recruiting, and legal), but the specific problems that a given startup 2 33 grapples with vary too much to offer support through standardized formats technical skills, who are also willing to such as workshops. The typical cy- work with young mobile entrepreneurs. cle of mentoring sessions should be between weekly and monthly. Mostly, An mLab or mHub generates most val- mentors and coaches are not expect- ue when it offers a portfolio of services ed to be high-level, well-established, that mutually reinforce each other. Most highly successful professionals. The mLabs and mHubs have experienced interviewed startups looked for depth that innovators benefit the most if and familiarity with their businesses’ they have the choice between vari- intricacies more than for big names ous means of support. For instance, and lofty visions. Yet, in the countries events can help an mLab to generate where mLabs and mHubs operate, it is a talent pipeline, while they give new- usually hard to find good mentors, even comers a chance to find out about the for payment. In particular, it has been a mLab’s offerings without the need to challenge to recruit competent subject commit to a lengthy support program. matter experts that know a given sector Incubation, on the other hand, is or industry, irrespective of business and more suitable for startup teams that The role of innovation competitions Innovation competitions for mobile app and ICT entrepreneurs have gained in popularity in recent years. They aim to provide opportunities especially for young innovators and technologists, for instance, offering cash, device, or incubation prizes. However, in places where ad hoc innovation competitions proliferated, they sometimes led to only temporary or no effects on innovative and entrepreneurial activity. For instance, competition fatigue and competition hopping are two worrying phenomena that have led those technologists to self-select for participation that are not able or willing to take forward and scale their products. Yet, with the right incentive structures in place, innovation competitions have shown to be useful to identify talent, become lightweight tools for startup acceleration, and extend tech innovation hubs’ reach beyond their physical space. In the experience of mLabs, mHubs, and infoDev, innovation competitions ought to be integrated with more continuous incubation support. infoDev has Box 4: The Role of attempted to capture this approach in its “From Mind to Market” approach to Innovation global innovation competitions, as well as in the implementation of the m2Work Competitions challenge and hackathon, the mAgri challenge, and the VentureOut challenge. More information on the From Mind to Market approach and innovation competitions can be found at http://www.infodev.org/m2Work, http://www.infodev.org/mAgri, http://ventureoutchallenge.org/, http://blogs.worldbank.org/psd/evolution-startup-competitions-case-pivot-east, and http://blogs. worldbank.org/psd/oh-no-not-another-apps-challenge. 34 such as hackathons, need substantial follow-up to achieve any impact on entrepreneurship. Particular emphasis should be given to coaching and men- Trainees were toring, as well as partnership and in- more likely to vestor contact brokerage. Competitions found startups also need to have strong differentiators if the training depending on local needs. If designed strongly emphasized well, competitions can be a viable means to interconnect mLabs and entrepreneurship. mHubs, and give early-stage innova- tors an opportunity for international exposure. have already shown some progress. Training programs are likely to have posi- Similarly, startup competitions and tive effects on the local ecosystem, but they “lightweight” accelerator models were do not appear as the most efficient means often seen as complementary rather for direct startup creation. All mLabs than a substitute for traditional incu- and several mHubs have implemented bation services. By providing a diverse training programs, ranging from boot- set of activities, mLabs and mHubs camps and workshops over few days can cater towards more diverse bene- to in-depth training over a whole year. ficiary needs, but also establish more These programs were aimed to tackle innovative and versatile partnership significant skill shortages that are im- models. It has also become apparent portant ecosystem gaps. Training was that ad hoc activities such as events generally well received by students and and competitions need to be held reg- hundreds of apps were created in the ularly to generate and sustain a brand process. However, startups rarely found and momentum within the community. trainees from mLab and mHub pro- Innovation competitions lead to better grams to be employable, as fledgling results for startup creation if they are em- companies mostly look to first recruit a bedded in continuous support, including core team of experienced and seasoned follow-up mentoring and contact broker- people. Often established technology ing with potential partners and investors. businesses actually benefitted more While competitions have not proven to from skills development programs, as replace in-depth support, they have many trainees went on to seek employ- become tools for “lightweight” acceler- ment with them. Trainees were more ation and extended the reach of mLabs likely to found startups if the training and mHubs beyond physical spaces. strongly emphasized entrepreneurship, mLabs and mHubs have learned that for example, by mandating the founding the most effort should be invested in the of a company or bringing an app to “before” and “after” of the actual com- the market as a course requirement. petition event (see box 4). Especially Still, the few startups founded in such 35 ideation and prototyping competitions, a manner were rarely sustainable. Training is a strong focus for mobile Strategy technology companies providing devic- Prioritizing value-maximizing activities es and operating systems, as it increas- has been a major challenge for mLab and es platform-specific coding skills and mHub managers as they have faced heavy the sheer number of available apps. resource and time constraints. Managers Also local governments found training are often overwhelmed by day-to-day to be important, given its contribution responsibilities and struggle to com- to developing the local labor force. plete just the grant deliverables. It is hard to pick correctly from the seem- mLabs and mHubs add substantial value ingly endless possibilities and opportu- for entrepreneurs and startups through nities to operate and partner, and they brokering relationships with clients, are usually understaffed. Mostly, mLab funders, and partners. mLab and mHub managers are supposed to assume too managers believe that their partner- many different roles, and rarely are they ship network should be cast as widely highly qualified for all of them. There is as possible, including linkages into a sense that the budget that was avail- non-tech sectors and international able under previous infoDev grants was connections. Partners have either not sufficient to attract known “heavy- added value for entrepreneurs directly weights” in the ecosystem, who would by offering innovation prizes, sponsor- find it easier to use their own resources, ships, mentoring, and in-kind contribu- experience, and networks to boost an tions, or they have supported startups mLab’s effectiveness. One proposal is to via the mLab’s or mHub’s channels, split the current mLab manager‘s role helping the mobile entrepreneurship in two: an external manager could take enablers with funding, branding, ex- on business development, partnership pertise, market information, and more. building, and external representation, mLabs have assumed various interme- while an internal manager would be diary roles; often they have functioned responsible for project management, as the trusted and recognized point of administration, and day-to-day support contact for large and risk-averse orga- of incubatees. One opportunity that nizations, thereby brokering business managers agree they missed is net- for incubatee startups. However, this working with and learning from other has so far mostly happened serendip- mLab and mHub managers, but they itously. mLabs and mHubs are only have usually had more urgent priorities beginning to formulate methodologies and see facilitation of this exchange as for matchmaking of startups and part- a task for infoDev. Managers also look ners, and they have not yet implement- to infoDev for more guidance and shar- ed specific relationship management ing of best practice for implementation strategies that startups could build and strategy (see Lessons Learned and upon. Startups and entrepreneurs Future Directions for infoDev). often feel that there is tremendous un- tapped potential in mLabs’ and mHubs’ It is important for mLab and mHub man- networks, including via infoDev and the agers to engage with the developer and 36 World Bank. startup community, but mLabs need to evaluate the effectiveness of community or mHub, they must find activities read- building for startup creation and mitigate ily accessible and affordable, including the risk of distraction. Managers found travel time and cost. Especially if de- that they have to immerse themselves mand for an mLab’s services is much in the community and earn its trust higher than its capacity, colocation with and respect. Inactivity and slowness other support organizations (such as as well as too strong affiliation with other technology innovation hubs, incu- government can lead to mistrust and bators, training centers, research labo- skepticism among members of the ratories, and universities) will improve community. mHubs’ experience shows the value proposition for entrepre- that it is important for community neurs. An mLab will benefit from close building to find the right combination coordination with colocated support of deep support for community cham- organizations, and clear differentiators pions and broad outreach to commu- and division of roles are paramount. nity members; this can be achieved In turn, it is hardly possible to create most effectively by mixing face-to-face a cluster of support organizations at a events with interactive online commu- location where mobile entrepreneurs nication. In turn, mLabs need to find do not usually convene. Science parks efficient ways to engage with existing removed from urban and student cen- communities if they are to maintain ters are rarely a preferred location for focus on direct startup support. mLabs the kinds of entrepreneurs that mLabs found that tapping into existing com- serve. Colocation with tech community munities or building niche communi- spaces (including open desk and co- ties provided more value to incubatees creation spaces) can help mLabs and and other entrepreneur clients. Where mHubs to attract interest from various communities were not readily available, stakeholder groups and enable seren- mLabs engaged more selectively, since dipitous exchanges and partnership basic community building rarely had building. However, mLabs also need to palpable effects on startup creation in make sure that its client entrepreneurs the short run. Ideation and networking and startups continuously venture out events were sometimes irrelevant or to explore customer needs and build even detrimental for incubatee startups partnerships with nontechnologists. that needed to focus on product devel- opment or exposure to customers and mLabs and mHubs benefit from building partners that typically do not engage in partnerships with a wide variety of stake- community events. holders that have a (partially) overlapping interest in entrepreneurship and startup An mLab’s location must be readily ac- support. mLabs and mHubs followed cessible, and colocation with other sup- an inclusive partnership building port organizations can provide the most approach, never focusing on a single value for developers and entrepreneurs. mobile platform or technology. mHubs’ If a community of developers, entre- and mLabs’ goals and incentives have preneurs, industry professionals, and been shown to overlap in many aspects investors are to form around the mLab with interests of private-sector tech 37 companies, entrepreneurship support turned into a strategic approach. For in- organizations, local governments, stance, such an approach would require training and education institutions, an alignment of incubation cycles with and so forth—and mLabs and mHubs startup competitions timelines, and it can become the nexus of these varied can be hard to achieve a smooth tran- stakeholder groups. While at times it sition between the competition process has been challenging to balance con- and sustained incubation. Similarly, flicting interests, mLabs and mHubs no mLab has sufficient tracking and have by and large benefitted from their reporting for startups in place for the neutrality, using it to position them- post-incubation and acceleration peri- selves as conveners that build the eco- od. All acknowledge that this can be- system as a whole, instead of dividing it come an important issue (for instance, up in a competitive fashion. for benefitting from the startups’ potential success, showcasing results, Events and competitions can be useful engaging alumni for coaching and men- tools to build partnerships and to select toring), but none of the mLabs has had incubatees. They provide partners with the resources or prioritized this task an opportunity to engage on a specific enough to fully address it. project, with a specific and limited framing. The risk of the approach is mLabs have struggled to achieve a strong that partners increase their expecta- regional footprint beyond their base coun- tions towards concrete deliverables try and city. The setup of local operations beyond branding over time. Moreover, was enough to consume most of the four a focus on startups and entrepreneurs mLabs’ resources. Stronger regional can get lost if events are dominated by engagement was achieved through sponsors that have a narrower agenda. startup competitions where submis- Competitions have also proven to be a sions from the region were accepted. useful channel for incubatee selection, However, providing needed hands-on as they highlight winners’ and finalists’ support for remote participants has potential and provide them with expo- been a challenge and can probably only sure and resources. be achieved through strong partner- ships with local technology innovation mLab managers are aware that selection hubs and incubators in other countries, of incubatees and tracking them after or through more sophisticated virtual they leave the program is crucial, but they incubation programs. have sometimes fallen short of doing this adequately. mLab managers are only Sustainability and Financial Planning beginning to set up explicit methods mLabs and mHubs have followed differ- to select incubatees that can benefit ent strategies to work towards financial the most from their services and that sustainability; activity-specific funding from have high-growth potential. Innovation tech partners and, for mLab Southern Africa and startup competitions have proven and mLab ECA, funding from provincial and useful for selecting incubatees, but national governments have been the main 38 it is unclear if and how this should be sources of income. mLabs and mHubs have not followed a single model to Overall, mLabs and mHubs found pursue sustainability, but rather relied that it was mostly possible to sustain on partners that, in any given local events, competitions, and training context, were willing to contribute and with sponsorship from private-sector would not overly dilute the mLabs’ organizations. Innovation competitions goals. Importantly, mLabs and mHubs can have substantial brand value for have learned that they should focus on ecosystem stakeholders, especially for brand building without requesting large large tech companies. The contribution financial contributions from partners will depend on the relevance and size in the first months and years, before of the target market for the sponsor, as an established brand helps to gener- well as good design and reach of the ate larger contributions and provides competition.23 Training is particularly stronger bargaining power towards relevant for large tech partners with startups when negotiating the terms of a developer outreach mandate (device future success sharing. For local gov- and operating system companies, ernments, training can be complemen- chip manufacturers such as Intel and tary to university curricula, seeing the Qualcomm, and some mobile network mLab as a quasi-polytechnic university. providers). If an mLab runs the train- Occasionally, this also implies a readi- ing as part of its core programming, ness to subsidize this kind of training. tensions can arise concerning the right mLabs and mHubs were successful in degree of a tech partner’s influence on securing financial contributions from pri- the training curriculum. Sometimes vate-sector partners primarily for events, tech partners opt for a more straight- competitions, and training, while mLab forward setup and simply contract the East Africa has also been successful in mLab to implement an explicitly plat- raising sponsorships for its core program. form-specific training. 39 Only in few cases were mLabs able to Startups and entrepreneurs are usually attract core funding that would cover open to paying for services or success the cost of overhead, infrastructure, sharing with mLabs, but they expect a and incubation from the private sector high-value service and are uncertain (for mLab East Africa); major contribu- about the appropriate procedure for tions instead came from government far-reaching contractual agreements. agencies (for mLab Southern Africa and Entrepreneurs are most prepared to mLab ECA). In absolute terms, govern- give back to the mLab that supports ment agencies have made the largest them, especially if they strongly identify coinvestments in mLabs. Especially with it. Worries and uncertainties con- in cases where the startup space is cerning equity dilution and the risk of too less advanced, tech partners (such as low valuations are widespread. While device and operating system providers some entrepreneurs strongly favor an and chip companies) can be more equity mechanism—assuming that this hesitant to invest in the ecosystem will create an incentive for the mLab to and have smaller budgets for software continue to support them—others are developer outreach and ecosystem wary and uncertain about the long-term building. Tech partners usually focus consequences; these entrepreneurs on the downstream of app innovation; tend to prefer rent or fee models. Some their key performance indicators of the entrepreneurs open to giving up revolve around app publications or equity do not find mLabs’ current value trained developers. Only where tech proposition compelling enough to give partners had larger internal budgets up a significant share without an addi- in more mature ecosystems did they tional financial investment on top of the shift focus somewhat more on the up- in-kind support they receive. Startups stream (such as app market traction, usually demand that success-sharing app startup creation). While partner- agreements be negotiated case by case. ship building with startups at events In turn, mLab managers are concerned and competitions is a goal for some about their lack of valuation and invest- tech partners, this is mostly secondary ment skills and, for simplicity’s sake, to brand building. Other private-sec- they tend to offer the same (or only a tor parties that would typically have an small set of) standardized contracts interest in startup creation (investors, specifying success-sharing conditions serial entrepreneurs) have not yet for royalty or equity sharing models to considered substantial direct contri- all client startups. butions to mLabs and mHubs given the client startups’ insufficient invest- At the end of the infoDev grant period ment readiness. mHubs are often able in June 2013, several mHubs and most to run Mobile Monday24 chapters and mLabs, despite progress in attracting rev- similar community activities continu- enue from services to private-sector part- ously without core funding, as small ners and entrepreneurs, project financing irregular contributions from many dif- gaps over the coming months and years. ferent sponsors usually add up to sus- No mLab or mHub is currently able 40 tain monthly to quarterly meetings. to finance its operations as a service provider, that is, only from direct returns by requesting startups to give up equity from customers paying for a specific stakes in exchange for incubation sup- service. mLabs have diversified their port or seed funding. It remains to be revenue streams significantly but the seen whether mLabs can ever be prof- short implementation timeline did not itable should they continue to focus on allow them to generate enough income turning talent into startups (see Should to fund operations independently of do- mLabs Focus on Ecosystem Building nors, government, and sponsors. Even or Startup Acceleration? and At Which including sponsorships, mLabs are not Innovation and Startup Stage Can yet able to cover the costs of their ac- mLabs Add the Most Value?). tivity portfolio from private-sector con- tributions alone (such as sponsorships, Based on the experience of the past service fees, and so on). Only mLab two years, infoDev estimates that, with Southern Africa (which secured addi- mLabs’ current focus on ecosystem tional funding from provincial and na- building and support for idea stage en- tional government) and several mHubs terprises, initial donor financing should (which run small, mainly event-based extend at least over the next six to ten operations) are guaranteed to sustain years. While many income sources are their current level of activity. The other imaginable for mLabs, each is tied to three mLabs are now looking for bridge significant tradeoffs and limitations funding including from public sources (see figure 7). Moreover, it is hard for (donors and local governments) while mLabs to monetize the service value slowly transitioning to revenue genera- of ecosystem building (for instance, tion from client startups, in particular, outreach to innovators at the ideation Figure 7: Royalties from startups Simplified Framework of Tradeoffs for an Equity liquidations mLab’s Potential Income Streams Donor & government subsidies for Short term availability entrepeneurship skills development of revenue? Potential share of On-demand technical training total mLab budget? Potential share of private Fees and sponsorships for sector contribution? events and competitions Focus on startup Management fees for outsourcing, creation? deal brokerage (“clearing house”) Donor and government subsidies for incubation and acceleration 0 1 2 3 41 Note: 1: Low, 2: Medium, 3: High. An ideal revenue stream would score “3” for all four decision factors. stage, training, community building, embrace the power of startup and tech and multistakeholder networking communities. It is helpful if managers and partnership brokerage) through are entrepreneurs themselves and if direct financial contributions from cli- they have established relationships ents. Instead, it appears that services with tech partners, investors, and other directed at ecosystem building will industry players. They need to inspire require subsidies aimed at long-term the trust of the ecosystem stakeholders impact and systemic changes in the that they serve, be passionate about enabling environment for innovation, building successful enterprises, under- with donors and local governments as stand social inclusion and development the typical stakeholder groups that can goals, and be nimble and ready to act provide such funding. If mLabs are to upon fast-changing trends in mobile maintain a strong focus on ecosystem markets. If no single candidate can building, they are likely to require a satisfy all the roles and responsibilities, much longer timeframe to reach fi- the mLab manager’s position can be nancial sustainability than the CSBKE divided into an outward-facing, repre- program allowed for; six to ten years sentative, and networking role and an seem more realistic than two or three inward-facing, administrative one. years. Extrapolating the revenue po- tential that has been uncovered by the An mLab consortium should be led by four existing mLabs, ecosystem build- a purpose-driven organization that has ing activities could become self-sus- an inherent interest in supporting the tainable without government or donor startup ecosystem. As with all consortia, contributions over the long run, for the goals and incentives of an mLab instance, through sponsorships, com- consortium’s member organizations munity membership or training fees, need to be aligned with the goals and contributions from community alumni, incentives of the mLab itself. But in and, to a lesser extent, through returns particular, mLabs need to be immersed from royalties and equity liquidations in the tech startup culture, as they rely from the few exceptionally successful on openness to and engagement with startups that have been incubated. innovation ecosystems and grass- roots tech and startup communities. Governance, Leadership, Accordingly, lead organizations in mLab and Consortia consortia need to have an inherent Identifying the right mLab or mHub man- interest and experience in tech entre- ager is critical for success. infoDev pro- preneurship support and community vides toolkits, materials and supervision, building, so that incentives are aligned but it cannot and should not micromanage and trust can be built. While, in theory, local implementation. Thus, mLab and a spectrum of organizations could fulfill mHub managers have great responsi- the consortium lead role, in most cases, bility to reach out to local stakeholders only a few organizations in an innova- and to communicate the value of their tion ecosystem will be able to strike the services. They need to believe in an right balance between private-sector 42 inclusive approach to partnerships and orientation and an ecosystem-oriented mission and purpose: a purely prof- and time are needed to prepare a it-driven organization might be led to high-quality proposal, manage other abandon activities with positive eco- consortium organizations, and handle system impact if there is no immediate the administrative requirements of en- profit potential, while an organization gaging with the World Bank. Especially that is detached from the private-sec- for small grassroots organizations (for tor and startup communities is unlikely instance, NGOs and community orga- to be a suitable choice. nizers), compensation—either through a management fee or through direct Consortia that are led by govern- (contractual) payment for service con- ment-linked entities or universities often tributions—should be openly discussed suffer from bureaucratic requirements and specified at the outset. Ideally, the that can conflict with the mission and lead consortium organization should mode of operation required for an effec- have surplus resources that can be tive mLab. mLabs need to be nimble, made available to the mLab temporarily, market-oriented entities. They cannot enabling the mLab to operate smoothly afford to be held up by bureaucratic and take opportunities quickly when roadblocks if they are to support en- grant disbursements are pending. The trepreneurs effectively and maintain need for institutional capacity has to be the trust of grassroots communities. weighed against the benefits of finding This requirement might be at odds with a purpose-driven grassroots organiza- lead consortium organizations that rely tion to lead the consortium; infoDev and on bureaucracies and intricate deci- donors have to carefully balance the sion-making structures, in particular tradeoffs and determine risk mitigation government agencies, universities, strategies. and large corporations. The mLab Each consortium organization needs to should not depend on potentially slow ensure internal stability and long-term decision making and political biases, backing of the mLab and its vision; risks of and the mLab manager needs to be internal challenges and bottlenecks have given the maximum possible executive to be mitigated early on. Organizations independence. Government agencies from traditional business sectors or should restrict their influence and fo- government can add value to a con- cus on securing additional funding and sortium, even if, for these actors, the high-level partnerships; they should startup culture and its implications for not interfere with the daily operation of an mLab’s strategic decisions are often an mLab, as this will most likely stall harder to grasp. Usually, an mLab re- progress and put the involvement of the lies on strong champions inside these startup community at risk. types of consortium partners. If a con- The consortium leader needs to have sortium member is faced with frequent sufficient institutional capacity and set or unpredictable management changes clear expectations and roles for its part- over which the mLab loses its champi- ners. mLab consortium leaders need ons, this will endanger the mLab’s op- erations. Each consortium organization 43 to be aware that substantial resources needs to ensure that its own decision formalize the basic rules of engage- making is in sync with the consortium. ment and their commitment early on For instance, hierarchies and potential- in a memorandum of understanding or ly slow approval processes in govern- contract. Any mLab consortium should ment and academic institutions need to agree on lean decision-making struc- be accounted for and militated against. tures, as nimbleness is critical for the mLab to succeed. Consortium member organizations are important channels for resources and If consortium organizations directly partnerships for an mLab; the organiza- contribute to the implementation of the tions ought to complement each other in mLab’s activities, the consortium needs this respect. Ideally, goals for the mLab to be able to manage potential conflicts and consortium organizations should of interest. It is generally acceptable be aligned by each consortium member if organizations join the consortium contributing to a vital function of the to contribute specific service compo- mLab in relation to the local innovation nents to the mLab, even if they expect ecosystem. For example, consortium financial compensation. This setup organizations could cover topics such as can generate strong commitment to funding for the mLab, startup commu- make the activity a success. However, nity access, mentoring and coaching, this constellation often reveals po- mobile technology expertise, access tential conflicts of interest: in simple terms, the organization in question to investors, training, and so forth. The has both a supervisory function as lead organization in particular should an mLab board member and also a be well networked and have a strong service delivery agreement with the reputation. All consortium organiza- mLab. For instance, the mLab could tions should be prepared to contribute be kept from canceling an activity regularly and throughout the whole du- that a consortium member imple- ration of the mLab project. Ideally, the ments even if the activity has become mLab management should continuous- irrelevant for the mLab’s own goals. ly update the consortium (for instance, The consortium has to be set up in a through virtual means) and seek feed- way so that such potential conflicts back about once per month. infoDev’s of interest are manageable and that belief in a grassroots-driven approach consortium members can be exclud- implies that it engages only as much ed if a given conflict becomes overly as necessary. In day-to-day implemen- strenuous. For instance, it is import- tation, the consortium organizations ant to formalize and specify the terms ought to be proactive and not depend on for any contribution from consortium “handholding.” If small grassroots or- organizations upfront, including the ganizations participate, the consortium grounds on which the contribution needs to account for typical challenges could end. Mutual trust, respect, and such as inadequate administrative transparency among consortium or- capacity and resource bottlenecks. ganizations are crucial to maneuver 44 The consortium organizations should any tensions and situations of conflict. Room for Debate: Open Questions for mLabs All existing mLabs and mHubs were pilot projects. infoDev selected a varied group of implementation partners across a diverse set of countries and cultures. Each manager was given only a basic idea of what an mHub or mLab could look like, and deliverables were kept rel- atively open-ended. infoDev encouraged managers to try out new approaches and push the boundaries of common incuba- tion and acceleration methods. This has meant that different busi- ness models were implemented. Most adaptations to the basic model were mandated by market and ecosystem needs, while others simply mirrored the visions and strategic decisions of in the startup accelerator model is to mLab and mHub managers. Hence, increase the client startups’ valuation, direct comparison of the pilots is lim- which then leads to higher potential ited, but they generated rich findings profits when exiting from equity posi- from the myriad of small implementa- tions (see figure 6 in section Services tion experiments. and Value Proposition). Rarely do pure accelerators have an incentive to pro- This section summarizes basic differ- vide training and community building ences in mLab and mHub managers’ beyond the point where this immedi- approaches, inconclusive findings, ately benefits the startup’s valuation as well as other open questions that and growth potential. mLabs and mHubs grapple with. Importantly, the choice between an eco- Should mLabs Focus on Ecosystem system builder and startup accelerator Building or Startup Acceleration? model has consequences for the type mLabs have been ecosystem builders of impact that is achieved. An ecosys- with elements of startup accelera- tem builder model is more inclusive: it tors (see section Services and Value gives opportunities to learn and benefit Proposition and Sustainability and to a greater number of individuals Financial Planning of mLabs and Hubs). and can facilitate a more diverse set Notably, “purer” startup accelerators of partnerships and exchanges. At the have taken a much narrower focus, same time, with more limited hands-on exchanging direct investments and in- support for each individual client and depth mentoring for equity positions in lower eligibility criteria, highly profit- 45 only a few startups. The primary goal able and fast-growing businesses will have a lower chance to be generated The question for current and future by an ecosystem builder model. What mLabs, then, is in which direction is more, the broad and often indirect they should orient themselves—more effects of an ecosystem builder mod- towards an ecosystem builder model el—even if ultimately substantial—will or more towards a startup accelerator also take more time to materialize and model. From infoDev’s perspective, this become obvious and measurable. question should be answered through an assessment of the largest potential It is also apparent that the integration for economic and private sector devel- of ecosystem builder elements into opment impact (or, more specifically, mLabs’ business models has conse- for the largest overall increase in suc- quences for viable paths towards finan- cessful, sustainable entrepreneurial cial sustainability. Given mLabs’ hybrid activity) as well as the viability of a character, they have pursued various given model in a particular market. revenue streams, diversifying their Importantly, infoDev and mLabs need income generation. However, these to consider several environmental revenue streams can sometimes be in conditions in local innovation eco- conflict with each other. For instance, systems, namely the landscape of the nimbleness, expertise, and rigor in existing innovation support institutions managing risk capital necessary to run (including technology innovation hubs), a successful startup accelerator can the most important ecosystem gaps, be at odds with due diligence and in- and potential cofunders’ priorities and clusiveness goals that can be required goals (see section Enhanced Scoping to deliver on funding for community Assessments and Ecosystem Mapping building and skills development. by infoDev). When examining the current orienta- tion and outcomes of mLabs, it is hard At Which Innovation and to envision them as profitable incu- Startup Stage Can mLabs Add bators or accelerators that are able the Most Value? to cover all their expenses through A strategic choice that will be decisive returns on their startups’ success—at for mLabs concerns the entrepre- least in the near future. In truth, the neur target group that promises the conditions in the markets in which highest potential for both impact and mLabs operate currently do not seem future revenue streams. This is relat- to afford mobile entrepreneurship ed to the question of whether mLabs enablers that achieve substantial im- should focus on ecosystem building or pact both as ecosystem builders and startup acceleration discussed pre- startup accelerators at the same time. viously, while homing in on the rele- For instance, overall, the majority of vance of entry criteria to receive mLab mLab-supported startups have shown support. So far, all mLabs have ad- limited investment readiness, even if dressed early stage innovation, where there are several exceptions of suc- structural support gaps exist in most cessful and fast-growing startups with developing and developed countries. 46 high valuations. But the existing mLabs have chosen to focus on different kinds of early-stage mLab East Africa, on the other hand, entrepreneurs, and this has had con- set a higher bar for selection for its sequences for mLabs’ strategic orien- incubation program, and several of the tation and results. more seasoned entrepreneurs that it hosted have found business success The comparison of mLab ECA and and international recognition soon af- mLab East Africa helps to highlight ter they joined the mLab.25 Even though some important differences. mLab mLab East Africa has not benefitted ECA is the best example of a focus directly from its incubatees’ success, it on “raw diamonds,” that is, talented was able to build its brand as an incu- young developers and entrepreneurs, bator of high-potential startups more mostly fresh university graduates, with quickly than mLab ECA. no prior entrepreneurial experience. mLab ECA also focused on team and Moreover, which is the right target motivation building in the initial stag- group will also depend on the availabili- es, without putting too much pressure ty of startup teams as well as the readi- on participants. While this approach ness, gaps, and needs of the ecosystem might have a higher marginal impact as a whole. Naturally, mLabs will pro- per supported entrepreneur, this will duce better results if they can build on a probably only become apparent over large talent pool and existing, well-or- time, as these clients might well go ganized communities of developers and through several failed startups before entrepreneurs. Wherever mLabs have they obtain the experience and intan- to spend significant time and resources gible skills needed to launch a sustain- for outreach, community engagement, able, fast-growing enterprise. and team building (as in the cases of mLab ECA, mLab East Asia, and mLab Southern Africa), they will generate lower results in the short term (see Results for mLabs and mHubs). Importantly, there is some indication that an approach that maximizes the economic impact can be at odds with one that creates revenue streams for mLabs beyond donor and government subsidies. In many contexts, inter- viewees stressed that a lack of skills is still the biggest barrier to high-growth entrepreneurship. At the same time, mLabs that focus on skills development rather than incubation and acceleration will hardly see successful startups emerge directly from their programs, so that it is harder to illustrate the 47 net effect on entrepreneurship of this approach. Moreover, as mLabs pursue independence from donor contribu- tions and expand the ways that they profit from equity sharing and royalty fees, they might be inclined to sup- port more advanced startups, which have a lower risk of failing and higher valuation potential. This would mean that mLabs become more akin to the startup accelerator model, while they fail to tackle the most severe gaps in the startup ecosystem. This illustrates how important better impact attribution will be for mLabs. At this point, it is unclear if mLab ECA’s long term approach (low selection requirements for incubation, in-depth intervention for a few innovators in- cluding a “safe space” for learning not been at the heart of their original and motivation) or mLab East Africa’s mandate. Social enterprise startups breadth and acceleration approach were able to contribute to mLabs’ (high selection criteria for incubation, brand building and recognition among, lightweight interventions for many for instance, donors, international de- velopment organizations, and impact fledgling entrepreneurs) will generate investors. For example, two of mLab the greater marginal effect on startup East Africa’s most widely noted start- success and economic activity (such ups are mFarm and Eneza Education, as additional revenues generated, which are both profit-oriented but also investments raised, taxes paid, and so role models for combining business on)—and, of course, these are but two success with positive social effects of many imaginable approaches. New through app usage. mLab East Asia has impact assessment methods should implemented a mobile hackathon for take a longitudinal perspective and UNICEF. Also the infoDev-run m2Work control for ecosystem conditions that project was an effort to push develop- are outside of a given mLab’s control in ers and entrepreneurs to tackle a topic order to determine its impact, and ulti- that was seen to have market potential, mately help mLabs to learn about and but also promised to yield app business improve their value propositions. models that would target mobile phone users at the base of the pyramid. Should mLabs Be Purely Market-Oriented or Accommodate These projects and mLabs’ affiliation Social App Enterprises? with infoDev and the World Bank as in- Some mLabs have supported social ternational development organizations 48 enterprises,26 even though this had has, at times, led to confusion among mLab stakeholders. They are frequently ultimately the scalability of supported unsure about the fundamental goals startups.28 Instead, these interviewees that mLabs pursue. In discordance to advocated that mLabs focus entirely infoDev’s and mLabs’ core mandate of on startups’ revenue and valuation fostering high-growth entrepreneurs, potential as the only way to identify some interview participants considered and nurture “gazelle startups” and mLabs primarily as subsidized incuba- “killer apps” that scale beyond nation- tors for social enterprises. al borders and lead to the maximum effect on private sector development. Of course, social and private sector de- Arguably, this would also make equity velopment goals do not have to be mu- sharing mechanisms more viable and tually exclusive. Some argue that pitting ultimately contribute to mLabs’ own “social” against “business” goals is a financial sustainability. flawed notion to begin with,27 especially in low-income countries where solving This has left mLabs with a difficult any user problem will often have some tradeoff: on the one hand, funding from positive impact on poverty alleviation. government, impact investors, and In particular, stakeholders of mLab other donors—which can be essential East Africa and mLab Southern Africa to cover an mLab’s core costs—is of- expressed the view that these mLabs ten explicitly or implicitly tied to social should motivate entrepreneurs to ad- impact and social entrepreneurship dress social problems, for instance, goals; on the other hand, it appears by developing business models that that an unequivocal orientation at address mobile users at the base of market and investor needs could help the pyramid. Mostly, those that favored mLabs’ success in bringing startups to mLabs pursuing a strategy that is at scale and support their own financial least in part guided by social develop- model in the long run. ment goals argued that mLabs should raise additional donor funds to be able Should mLabs Emphasize to subsidize support channels for so- Their Expertise in Mobile cial enterprise startups, even if their Technology, Entrepreneurship, businesses are less likely to achieve or Specific Subsectors? profitability and investment readiness At their core, mLabs and mHubs are under current market conditions. mobile technology-specific entrepre- neurship enablers. The underlying Yet, in particular venture capitalists idea is that mobile app and software and angel investors seemed skeptical markets work differently from tradi- about any strategy that would consider tional industry sectors and even from social impact goals next to pure start- other tech sectors, and that this has up success. Several interviewees ar- implications for how entrepreneurs and gued that embedding a donor-backed startups should be supported. subsidy for social enterprises into mLabs’ financial model raises the risk Yet, mLabs and mHubs have experi- of diluting their market orientation and enced time and time again that what 49 mobile app startups need is still id- developing countries are definitely dis- iosyncratic to any individual business tinct from other business communities. model—even for the mobile innovation Stakeholders emphasized the need for niche, there is no one-size-fits-all mLabs and mHubs to educate investors support methodology. mLab-supported and incumbent businesses about the startups often asked for typical incu- different culture of hackers and mobile bation services, such as one-on-one technologists, but also about differenc- mentoring for core business functions es of how mobile app startups work, for or in-depth tracking of milestones instance, with regard to innovation pro- and startup progress. In particular in cesses, monetization, valuation, financ- Vietnam, but also in other countries ing, and so forth. This would also imply where mLabs are active, interviewed that mentors need “mobile”-specific stakeholders pointed out that the lack expertise to better help startups. The of entrepreneurship and business motto “mobile is what we do” has been skills are a much greater constraint for embraced by mLab Southern Africa: startup sustainability than insufficient it recently embarked on a strategy to technical skills. This could mean that complement established tech and gen- mLabs should lessen their focus on eral business incubators with a mobile “mobile” and technical expertise, and tech pillar, through, what it calls, a rather perfect technology-agnostic in- “vertical model” (see case study on cubation and acceleration services. mLab Southern Africa). Among other things, this brings the advantage that On the other hand, startup communities the mLab can build a reputation as a are often born out of tech communities, one-stop shop for all stakeholders that 50 and the mobile tech communities of are interested in mobile app solutions or want to reach app developers and services to incubatee peers and the mobile technologists. mLab itself, resulting in an mLab-in- ternal market and learning platform. Yet another position holds that, most The mLab identified functions such importantly, tech startups need to be as app development, graphic and web grounded in the customers’ reality and design, technical training, digital me- improve on incumbent ways of serving dia and gamification, cloud services customer needs, which can be aided and mobile app infrastructure, as best by engaging non-technical sector well as community outreach as areas and subject-matter experts. The pro- that are relevant to many mobile tech ponents of this opinion usually suggest startups. Although not all interviewed that a focus on “mobile” and a close stakeholders said that peer-to-peer connection to tech communities is vital learning was essential for clients of for mLabs, but that they should build mLab Southern Africa, others re- topical sub-communities and urge counted anecdotes of fruitful collabo- startups to venture out of both the tech rations. Maybe even more importantly, and entrepreneurship bubbles in which the mLab itself was able to utilize the they often find themselves, in order to incubatees’ assets, either by directly get closer to their customers. contracting them for internal projects or by subcontracting them to service Ideally, mLabs unite the best of all mLab’s external partners. mLab worlds, and cater their support to Southern Africa is still in the experi- startups depending on their needs and mentation phase, but so far the com- business model gaps. However, mostly plementary portfolio of startups fits mLabs will have to make a judgment well into its differentiation approach call on what is the right positioning based on “mobile verticals” (see case for the market in which they operate. study on mLab Southern Africa). At least, mLab managers need to be aware of the risks and opportunities Other interviewees were more skepti- of each approach, and continuously cal. They highlighted that incubatees identify which one is best able to fill should be selected based on clear and (potentially shifting) support gaps in universal requirements. According to the ecosystem. this perspective, the mLab’s success is closely coupled with the success that Should mLabs Build a a startup has on its own long after it Portfolio of Incubatees with has left the mLab’s support program, Complementary Assets? as it affects both the mLab’s returns Interviewees also disagreed on from equity liquidations and its brand. whether mLabs should consciously Constructing a portfolio of comple- build a portfolio of incubatee startups mentary incubatee startups, in the that complement each other. The de- opinion of these interviewees, would bate was inspired by mLab Southern necessarily dilute competitive selec- Africa, which selected incubatees in tion and limit the selected startups’ a way so that they can provide their entrepreneurial freedom. 51 Should mLabs Make Direct to limit themselves strictly to bridging Equity Investments or Broker the so-called market failure but they Investor Contacts? should not distort the venture capital Interviewees also disagreed whether and angel investor space. In particular mLabs should make direct financial in hyped and fast-growing ecosystems investments in startups in exchange for such as Nairobi, it can be hard for equity shares. Many interviewed entre- mLabs and other entrepreneurship preneurs pointed out that a combina- support organizations to determine tion of seed investment and incubation up to which point promising startups support would be the biggest help for should receive privileged access to their business, and that this support capital and at which point startups need mix would also heighten their readiness to be exposed to market-based demand to share equity with the mLab. This is and monetization constraints in order in line with infoDev’s experience that to be able to mature and advance. critical financing gaps for early-stage innovators exist in the pre-seed invest- An additional challenge is the mLabs’ ment stage.29 currently lacking capacity to make in- vestment decisions. Many interviewed But other stakeholders requested that startups requested that any equity mLabs only guide startups to the point sharing agreement be negotiated where they can attract investments un- case by case, but mLab managers der market conditions. mLabs should and staff hardly have the resources strengthen the startups’ investment and experience to assess the fledgling readiness—for instance, by focusing startups’ valuation correctly. Moreover, on monetization strategies and mar- mLab managers’ own incentives would ket traction, or by teaching pitching have to be adjusted, for instance, by skills—and then broker contacts to compensating them in cases where private investors, without making a startups raise additional capital with direct investment. This argument holds their help. Hence, several stakeholders that mLabs, as partially donor and worried that, under current conditions, 52 government subsidized actors, need mLabs would make suboptimal use of a seed fund, unless, as a minimum, effective, and adjust and refine their an investment expert can be hired to business models. For this endeavor, it oversee it. Several mLab managers as will be key to strike the right balance well as some entrepreneurs advocated of in-depth scoping assessments and an alternative setup where mLabs offer high-level business planning for mo- a fixed take-it-or-leave-it equity propo- bile entrepreneurship enablers on the sition to limit the required negotiation one hand, and avoiding superimposing and management effort. This could also a set structure onto local partners or lead to a positive self-selection effect: if micromanaging their operations on an mLab’s blanket offer is calibrated in the other. line with its value proposition, startups that request better conditions might A natural fit for infoDev’s positioning is already be advanced enough to pursue enhanced innovation ecosystem map- investments from private angel inves- ping. In its current approach, infoDev tors or other channels. assesses the following factors in an ecosystem: Lessons Learned and Future • The innovation and entrepreneur- Directions for infoDev ship enabler landscape (including incubators, technology innovation With the CSBKE program, infoDev had hubs, accelerators, government an opportunity to push the envelope of funded programs, and tech commu- support interventions for mobile in- nity organizations and networks); novation in developing countries. Now • The availability of talent and that the first mLab and mHub pilots human capital; are concluding and the network is ex- • The current market size and panding further, it will be crucial for in- available monetization channels for foDev to do a better job at consistently mobile app startups (for instance, supporting local enablers and improve availability of credit cards or mobile overarching project management and payments, mobile advertising administration. This section discusses industry, and so on); the evaluation’s insights pertaining to these topics. Given its positioning as a • The access to finance landscape; trust-funded program within the World • The state of the ICT sector; Bank, the following lessons also bear • The overall regulatory, policy, and consequences for infoDev’s donors and business environment; and relevant units inside the Bank. • The overall ICT infrastructure.30 Finding the Right Mobile The assessment process involves sev- Entrepreneurship Enabler Model: eral country visits during which stake- Enhanced Scoping Assessments holder workshops and focus groups and Ecosystem Mapping are conducted, while stakeholder Going forward, infoDev will aim to involvement is increased over time to pick implementation contexts where codesign the envisioned mobile entre- mLabs and mHubs can be most 53 preneurship enabler. 54 Ecosystem Co-funder Figure 8: Innovation support gaps priorities Environmental landscape Factors and mLabs’ Strategic Positioning as Ecosystem Builders or Startup Accelerators mHubs mLabs Ecosystem Startup builder accelerator infoDev has begun to use this framework account the complex linkages between and transfer the knowledge from the ecosystem stakeholders. This will be mLab and mHub pilots under CSBKE critical to identify and anticipate eco- to new programs: in 2012, infoDev system gaps and stakeholder needs, agreed with the Canadian International as the only way to maximize the value Development Agency (CIDA)31 to roll addition of future mobile entrepreneur- out a Mobile Innovation component ship enablers. for several islands in the Caribbean More in-depth scoping assessments and over a seven-year timeline, in this case better innovation ecosystem mapping leaving more time for infoDev-led scop- also appear as the most viable approach ing and codesign as well as the local to answer the question of whether implementer’s path to sustainability. mLabs should focus on an ecosystem Similarly, in several countries of West builder or startup accelerator model (see Africa including Senegal and Nigeria, sections Services and Value Proposition, infoDev is currently conducting scoping Sustainability and Financial Planning of assessments with support from the mLabs and Hubs, and Should mLabs Swedish International Development Focus on Ecosystem Building or Startup Agency (SIDA). Acceleration?). Factors potentially in- While this approach is likely to mitigate fluencing the decision can be grouped several of the problems that the mLab into: the landscape of existing innova- and mHub pilots have faced, infoDev tion support institutions (including tech and others are pressed to find more innovation hubs and labs); the most sophisticated methods for ecosystem important ecosystem support gaps; and mapping, including quantitative as- potential cofunders’ priorities and goals 55 sessments and methods that take into (see figure 8). For instance, insufficient early-stage risk capital together with due to an initial lack of insight into the availability of investable startups could local innovation ecosystem or a limited mandate a startup accelerator model, understanding of what an ecosystem while a serious skill shortage or a lack support project should usually entail. of connectedness of stakeholders as the most important ecosystem barri- Most mLab managers and consor- ers would point towards an ecosystem tium members also agreed that builder model. Such an approach will the size of each mLab grant was help to configure mLabs’ and mHubs’ too small to make optimal use of unique value addition, one ecosystem the opportunities that mLabs were presented with. mLabs felt they did at a time. not have sufficient funding to build Program Framework: Timeline, up internal capacity or hire external Grant Size, Scalability, and Design help to adequately service the many Interviewees also had suggestions for stakeholder groups, and fully engage improvements in infoDev’s overall pro- with startups and other clients, as gram framework. The least contested well as partners, funders, and other innovation ecosystem stakeholders. finding was that mLabs need a much Especially the mLab managers’ time longer timeframe to set up an oper- was stretched thinly across the many ation that is smooth and sustainable, responsibilities that they were given, including a honed mission and goals, and potentially mLabs would have a consistent operational and partner- been more effective if the grant size ship model, results that show its full had allowed for recruiting two man- potential, and the ability to consistently agers with complementary roles. cover the cost for the portfolio of activ- ities. Within implementation periods of At the same time, both the short time- between one and two years, no mLab line and relatively small funding per was able to arrive at such a state. Well- mLab or mHub project were deliberate known success factors for incubation program design choices, with the goal programs include rigorous selection to enable a greater number of projects and follow-on tracking of startups, that would generate more and deeper which mLabs also have not been able insights. In fact, infoDev’s experience to complete to perfection given time with business incubators had shown and resource constraints. Finally, the that, at the minimum, three years are fact that mLabs are deeply intertwined needed if a new incubator is to be set with complex tech community dynam- up, and at least six years are usually ics and attributes of the local innovation required for it to reach financial sus- ecosystem often led to slow progress tainability. The two-year implementa- in the initial implementation period. tion periods under CSBKE were seen It was hard for some managers and rather as a seed-stage for the mLab or consortium organizations to correctly mHub before promising models would gauge what an mLab’s business mod- receive follow-on support for scale up, el could look like and which activities including funding, towards the end of 56 should be prioritized. This was mostly the grant period. It remains that there is a tension be- a fixed long-term implementation plan tween infoDev’s dedication to piloting on the grantee. At times, there might and learning (mandating short time- already be a local organization with a lines and small budgets) and the need strong vision or even implementation for consistency in the projects’ oper- plan for a mobile entrepreneurship ations and their scale up (requiring enabler concept. The many fruitful sufficient initial capacity and follow-on adjustments that mLabs were able to funding in the mid term). While mLabs make based on local conditions illus- in particular are currently undergoing trate the core advantage of infoDev’s a difficult transition after the end of the grassroots and demand-driven project grant period, infoDev has identified the design, that is, to leave the calibration importance of flexible follow-on fund- of business models to local partners ing based on a local enabler’s needs that have the trust, contact networks, and previous success. Future mLab and responsiveness that is required to grants can either have an extended roll out a successful ecosystem-orient- timeline to begin with, include (provi- ed project. Similarly, infoDev will aim to sional) matching funds from other par- be acutely aware of market needs, and ties, or be designed to leave room for be prepared to pull out of a country or additional fundraising before or imme- explore entirely new interventions if diately after implementation begins. given conditions can endanger the ba- infoDev’s newly established multidonor sic viability of an mLab. trust fund could become an important institutional step towards such more Grant Administration flexible arrangements. and Procurement Stakeholders as well as mLab and Moreover, for infoDev the identified challenges imply that it could cover at mHub managers pointed to substan- least the higher level business mod- tial issues that, they felt, came about eling and planning before a grantee due to arduous grant administration begins project implementation, which and procurement rules. Practically all can be based on more in-depth scoping mHub managers were overwhelmed assessments (see Finding the Right with reporting duties and other admin- Mobile Entrepreneurship Enabler istrative barriers, and mLabs at least Model: Enhanced Scoping Assessments felt that these tasks were a strain on and Ecosystem Mapping). It appears their commitment to implementation. that infoDev’s experience gathered from innovation ecosystem support around To be sure, grants are infoDev’s core the globe can translate into good aware- business and (except for short-term ness of how ecosystem dynamics play service contracts) they are the unique out and how mLabs can fill important way for infoDev to engage with its cli- gaps. Yet, infoDev will maintain a partic- ents. Yet, it is understandable that the ipatory program development process procedures that are tied to grant ad- and codesign its interventions with ministration can seem disproportionate local stakeholders, limiting its own role to grantees, especially in cases of small grants and when grantees have no prior 57 to conceptualization, without imposing experience with grant administration. (such as innovation competitions) could Sometimes, World Bank procurement be implemented through a “hub and guidelines entirely prohibited small or spoke” system, where a larger entity young organizations from tendering. absorbs the grant (or contract) and On the other hand, mLabs and mHubs subcontracts mLabs and mHubs in can often benefit more from their affil- small target countries or areas within iation with infoDev and the World Bank a larger country. infoDev can also take than grant administration costs them. a more active role during the grant Moreover, both grantees and infoDev preparation period, ensuring that bud- can learn from the pilot projects and gets follow a structure that complies streamline procedures to avoid at least with World Bank financial management some of the challenges. and procurement rules and discussing expectations for reporting results in For future grantees, the key lesson more depth. infoDev can improve on the is that they should account for the provision of good quality templates and administrative burden ahead of time. guidelines for these steps and include They should ensure that they earmark examples from past infoDev activities budgets for auditing, staff who handle that are more akin to an mLab or mHub procurement and financial reporting, manager’s project than regular World and staff who can collect monitoring Bank templates can be. In particular, and evaluation data and compile draft during the first reporting rounds, in- reports for the mLab or mHub man- foDev will be prepared to provide more ager’s review. Similarly, grant admin- detailed instructions and administrative istration, by its nature, takes time, as assistance to grantees. During scop- it involves checks and balances and ing assessment and other in-country coordination effort, so, to some extent, missions but also through webinars, grantees simply need to prepare for infoDev can also provide small training (seemingly) slow response times. In sessions on grant administration, again the same way, legal constraints can employing use cases that are more sim- sometimes prohibit infoDev from pro- viding grantees with time-sensitive ilar to what the mHub or mLab project information, especially in the early in question is expected look like. Lastly, stages of drafting grant agreements. infoDev can be more aware of the grant- Moreover, the World Bank is currently ees’ implementation priorities; for in- revising its policies and procedures stance, it could send frequent and early further and this can have implications reminders about reporting deadlines for infoDev’s way of working; grantees so that grantees do not risk prioritizing are encouraged to clarify requirements administrative tasks too late. as early as possible. Knowledge and Analytics through For infoDev, several valuable lessons Improved Monitoring and Evaluation emerged from the mLab and mHub Given mLabs’ and mHubs’ function as pilots implemented under CSBKE. pilot projects, infoDev claims a strong Importantly, mHubs as well as dis- mandate to conduct learning-oriented 58 tributed global and regional activities monitoring and evaluation. For mLabs - Services offered Services offered Services by one or occasionally by not offered more mLabs one mLab by any mLab and mHubs, the value that infoDev Figure 9: Venture Funding mLab Services can add stems from its ability to pull as Originally information from individual projects, Envisaged Business Training aggregate and compare it, and finally package the insights into applicable Mentoring knowledge products and toolkits. While mLabs and mHubs have most- Technical Training ly appreciated that infoDev remains “hands-off” with regard to project implementation, they expressed the - Testing & Certification view that it ought to be more active and nimble in feeding findings back to - Due Diligence & Capital Readiness the local level. Value Chain Partners The original generic mLab business plan32 and workbook that infoDev pro- Market Intelligence vided at the beginning of the program was described by most mLab managers Physical Space as a useful conceptual guide, but of lit- tle help for prioritization and day-to-day Technical Outsourcing decision making. The generic business plan seemed to highlight what is possi- Professional Services ble, but not what could realistically be achieved, especially in view of a short Competitions implementation timeframe and evolv- ing market conditions. Interviewees Networking Events mentioned that governance and ca- pacity issues often stood in the way of - Code Repository realizing what conceptually would have been ideal. The workbook was seen as Note: A checkmark indicates that the services too elaborate and specific to apply to were offered by one or more mLabs. A red the malleable and constantly evolving dash indicates that the service was offered only business modeling and planning that occasionally by one mLab. Venture Funding and Market Intelligence were not offered as an mLab managers have been conducting. explicit service line by any mLab. Moreover, mLab managers had to adapt to emerging und unexpected priorities of clients. For instance, mentoring - Evaluation, Business Analytics, turned out to be more important than and Toolkits expected, while testing and certification Overall, infoDev is determined to en- was less relevant than imagined. Vital hance its ongoing value as an analytics services such as direct venture fund- and evaluation provider for mLabs and ing and market intelligence were not mHubs. In particular, many insights offered because of capacity constraints about mLabs’ and mHubs’ potential and (see figure 9). 59 past success lie in detailed data of the supported startups and entrepreneurs, for mLabs and mHubs, Sustainability but mLabs and mHubs have not had the and Financial Planning, and At Which resources to collect such data contin- Innovation and Startup Stage Can uously and rigorously. While evaluation mLabs Add the Most Value?), mLabs fatigue and concerns about disclosure and mHubs can have clear opinions of financial data are common among on which ecosystem gaps are most client startups, infoDev could use its critical, but tackling them can be at analytical experience and global reach odds with the possibility of showcasing to compile and evaluate relevant data in impact and the mLab’s or mHub’s own a non-intrusive and efficient way. financial sustainability. While, concep- tually, it is clear that mLabs should infoDev reacted to mLabs’ and mHubs’ focus on the early stage of innovation evolving information demands by and high-growth entrepreneurs to have commissioning this report (focusing the greatest effect on economic de- on “what was” rather than “what could velopment, it is difficult to capture all be”), as well as upcoming toolkits on positive results of such an approach, business model design and business especially if startup success does not analytics for mLabs and mHubs that materialize in the short run and is in- are envisaged to be more customizable terdependent with complex innovation and applicable. Such commitment to ecosystem dynamics. Longitudinal improved analytical support for mLabs studies of supported entrepreneurs and mHubs is commendable, and wide and startups, ideally with counter- publication of knowledge products factuals and controlling for external could be of value for technology inno- ecosystem factors, are the only way to vation hub and incubation practitioners increase certainty about which inter- beyond infoDev’s network. However, ventions result in the largest impact, infoDev is aware that its own knowledge even if they are more expensive to demands, the demands of donors, and carry out. With such analysis, it could the knowledge gaps for local entrepre- be confirmed whether contributions neurship enablers do not overlap per- by donors and funders of mLabs and fectly. Even more importantly, infoDev mHubs are needed and effective, and will aim to ensure that monitoring and how much time they take to materialize evaluation techniques feed into short in palpable economic impact. feedback loops that enable infoDev and its stakeholders to continuously build, Impact assessments could also go measure, and learn.33 beyond effects on startup creation and economic development. First, - Holistic Impact Assessments infoDev could help measure and infoDev will also benefit from a stron- quantify the overall effect that mLabs ger effort to support more rigorous and and mHubs have on innovation eco- holistic impact assessments for mLabs systems, especially in view of recent and mHubs. As described in this report improvements of ecosystem mapping (see sections Why Evaluate Business and ecosystem quality assessment 60 Models of mLabs and mHubs?, Results methods (see Finding the Right Mobile Entrepreneurship Enabler Model: network it should focus on and which Enhanced Scoping Assessments and ones it should drop. By and large, the Ecosystem Mapping). Second, apps interviewed stakeholders advocated that lead to positive spillover effects for that infoDev strengthen its own team poverty alleviation add to the holistic resources and provide more assis- impact that mLabs and mHubs have, tance throughout the implementation but so far these effects remain unmea- period of an mLab or mHub project, sured. infoDev could help to capture but, of course, the largest possible the benefits (and risks) that users share of funds should flow to local experience from engaging with the app economies and implementations on innovations that client entrepreneurs the ground. The recent establishment of mLabs and mHubs produce. This of a multidonor trust fund for infoDev does not only include traditional ICT for will be a major step towards greater development sectors—such as mobile flexibility and nimbleness to invest agriculture, financial inclusion, health, resources in a way that they maximize education, or government—but also the positive effect on innovation eco- (with a broader notion of poverty allevi- systems in developing countries. ation in mind) entertainment apps and games, productivity and supply chain - Setup of an Exchange and optimization apps, or social media and Collaboration Platform Between interpersonal communication apps. in- mLabs and mHubs foDev has recently made a first step in Since their inception, mLabs and this direction: it commissioned an out- mHubs have hardly collaborated or come assessment of mLabs’ holistic exchanged knowledge with each oth- impact from the evaluation consultan- er. There was widespread agreement cy Global CAD, to be released in March that this implies that mLab and mHub 2014.34 Moreover, as a follow-on study managers are missing out on relevant to prior research on behavior of mobile knowledge and partnership oppor- users at the base of the pyramid, 35 in- tunities, in particular, those aimed foDev commissioned a database of 50 at global market extension for client apps in several African countries that startups. benefit those users and have created revenue. Interviewees and discussion partners from various backgrounds advocat- Continuous Global Network ed that infoDev create a platform for Facilitation and Technical Assistance mLabs and mHubs to interact. However, for mLabs and mHubs it was unclear for these stakeholders infoDev’s Mobile Innovation Program what exactly such a platform should as a whole is presented with a host of look like and how it should be managed promising pathways to strengthen the and financed. Some suggested a sim- global network of mLabs and mHubs. ple web-based platform with limited Yet, infoDev will need to prioritize intervention by infoDev, leaving it up to and carefully evaluate which of its mLabs and mHubs to define the rules 61 own activities aimed at convening the and intensity of the engagement. However, infoDev has already at- innovation hub associations, such as tempted to foster interaction between AfriLabs. While an mLab and mHub as- mLabs and mHubs with this kind of sociation certainly seems like a worth- “hands-off” approach. It invited mLab while project in the long run, current and mHub managers to engage with priorities, resources, and capacity— each other on the public, web-based both on infoDev’s side and on the part discussion board mLabs World, with- of mLabs and mHubs—might not favor out further facilitation. It turned out a separate formalized institution. that no active exchange occurred. The website was practically unused when - Global Activities Run by infoDev it was finally disabled in 2013. infoDev Instead of facilitating a continuous ex- also brought mLab and mHub manag- change and collaboration between lo- ers together annually for closed-door, cal enablers, infoDev supported mLabs face-to-face peer-learning workshops (and some mHubs) through temporary during large conferences, namely in- global activities. The core costs were foDev’s Global Forum36 and the Nokia- covered directly from infoDev funds, led Open Innovation Africa Summit.37 while the projects would leverage While these sessions were seen to existing mLab and mHub operations be stimulating and helpful, an annual for local implementation, outreach to meeting rhythm was not enough to developers and entrepreneurs, as well jumpstart continuous exchanges. as partnership building. More recent discussions led to the idea So far, infoDev has mainly used in- of creating an association of mLabs and novation competitions. The m2Work mHubs. The terms for the exchange project, with additional support from and eventual contributions would still the United Kingdom’s Department for be left to mLabs and mHubs to decide, International Development (DFID), but infoDev could play an active role piloted infoDev’s From Mind to Market in the setup process. Stakeholders methodology,38 which sources mobile agreed that there would have to be app ideas through crowdsourcing and core funding for several years to give guides the best ones towards incuba- the association time to clarify its role. tion and acceleration through mLabs. The ensuing mAgri Challenge39 (sup- At this point, it is difficult to assess if the ported by SIDA) and VentureOut40 chal- formalization of an association would be lenge followed a leaner process and a valuable endeavor for infoDev. mLabs focused on scalable business models have little spare time and resources and business internationalization that they could invest, which implies respectively. infoDev also used these the risk that their engagement will re- projects to push novel app business main limited both as contributors and models in thematic areas or sectors, service recipients. infoDev, mLabs, and gauging their potential for inclusive mHubs would also have to coordinate innovation: m2Work sought app start- with and differentiate themselves from ups that employ mobile microworkers 62 ongoing efforts to establish technology at the base of the pyramid; the mAgri Challenge aimed to identify scalable build business models based on the businesses that have the potential to findings of the study. Similarly, infoDev spread the positive impact of mobile is currently conducting virtual incu- agriculture apps to more farmers bation pilot projects with strong ana- and agricultural workers across Sub- lytical components in Vietnam44 (with Saharan Africa. the participation of Topica MSN, the mHub Vietnam) and East Africa45 (with Despite capacity and timing problems the participation of mLab East Africa). for some of the partnering mLabs and infoDev can build on such approach- mHubs, by and large these innovation es and collaborate with mLabs and competitions were seen to add value mHubs to release analytical products for local mobile entrepreneurship that contribute to local stakeholders’ enablers and their clients. It became understanding of market opportunities clear that infoDev-run global inno- or gaps in the ecosystem. vation competitions can combine the best of both local and global support if - Global Fundraising, Partnership they are well managed, provide mLab Building, and Contact Brokerage and mHub managers with adaptable toolkits for efficient local implemen- Many stakeholders demanded that tation, and address an actual need infoDev be more active in global fund- for entrepreneurs which mLabs and raising and partnership building for mHubs cannot satisfy on their own. the Mobile Innovation Program, for the infoDev can build on this experience ultimate benefit of mLabs and mHubs. and continue to implement selected, In view of the short implementation well-defined global activities that timeline under CSBKE and projected complement and leverage local entre- financing gaps over the coming years, preneurship support channels. most mLabs and mHubs looked to in- foDev to provide or facilitate follow-on Finally, market research based on the donor funding. Several interviewees participation of local mLab and mHub also expected infoDev to raise more stakeholders could become another mutual awareness and broker rela- value generator both for infoDev and tions between relevant World Bank the participating entrepreneurship en- units and mLabs, whenever this would ablers. In 2012, infoDev ran two studies provide branding, outreach, and fund- on mobile usage at the base of the ing opportunities. pyramid, in Kenya41 and South Africa,42 as well as an overview report.43 For Many interviewees expressed a positive both case studies, the respective view on public-private partnerships mLabs participated in the research in and advocated, for instance, that oth- that they gave in-depth insights into er global tech companies be involved local mobile innovation ecosystems or in infoDev’s network-level planning. provided access to entrepreneurs and Corporations such as Qualcomm, other stakeholders. In Kenya, addition- Google, Intel, BlackBerry, Microsoft, al workshops are envisaged to help and others have programs in place that 63 mobile developers and entrepreneurs support tech entrepreneurs across the globe, and infoDev’s role was seen to project-based partnerships, in partic- include general outreach to these units, ular, for innovation competitions and as well as brokerage of direct contacts conferences, but has not struck another for mLab and mHub managers. global partnership with a private-sec- tor actor following the agreements So far, the CSBKE program’s most with Nokia and AppCampus. It limited important private-sector partner has itself to partnerships where no funds been Nokia. The company’s represen- or long-term commitments had to be tatives contributed expertise to early exchanged, leaving more substantial conversations between the government agreements up to mLabs and mHubs. of Finland and infoDev, helped design It stands to reason that infoDev will conceptual material for mLabs, and consider a stronger emphasis on global provided substantial implementation fundraising and partnership building and governance support for mLabs for the Mobile Innovation Program. in South Africa, Vietnam, and Kenya. AppCampus could be just the first of Later, Nokia helped to set up privileged many extended partnership agree- access to funding and support for mLab ments. While it takes time and effort entrepreneurs from the global app to achieve concrete contributions from startup accelerator AppCampus.46 Yet, partners, infoDev’s expertise, conven- the partnership also brought to light ing power, and impartiality could make challenges that were reflected in early it possible to crowd together venture discussions with other potential global and app developer-oriented units of partners: Nokia did not have the same global tech companies, impact inves- level of interest in all mLab markets (for tors and venture funds with a focus on instance, it did not partner with mLab developing countries, entrepreneur- ECA); it preferred supporting common ship-oriented foundations, interna- developer outreach channels such tional organizations active in mobile as technical training and prototyping technology, and so forth.47 competitions over incubation and direct startup creation; and it faced other in- - Responsive and Adaptive Change ternal challenges such as constrained and Conflict Management resources for coordination across mLabs and mHubs have had mixed countries and budget limits. expectations regarding infoDev’s role as an active facilitator of local infoDev also encountered its own in- operations. For instance, mLab East ternal barriers to building more and Asia asked infoDev to engage more deeper global partnerships. infoDev actively while mLab East Africa explic- found that the formalization of long- itly commended infoDev for remaining term partnerships would result in sub- detached from local implementation. stantial transaction costs, for instance, mLab ECA and mLab Southern Africa to ensure legal viability and alignment did not expect infoDev to get involved with grant and trust fund rules. As a re- in matters of project implementation, sult, infoDev has so far opted to engage but wanted it to communicate its 64 in several informal, lightweight, and mission more clearly to stakeholders and clients, and to be more active in on milestones, and intervene actively leveraging the World Bank brand and at the local level when the grantee is brokering partnerships. undergoing a change process or other governance and management conflicts. A more consistent finding is that in- foDev had difficulties in anticipating Evaluations of the collected data com- and resolving unexpected implemen- bined with outcomes of infoDev’s eco- tation bottlenecks and conflicts. For system mapping and research agenda example, infoDev found that it could can also be packaged into formats that have applied itself more to change are useful and accessible for mLab and management when there were sudden mHub managers. infoDev is currently personnel changes on the grantee’s planning to develop customizable and side, as in the cases of mLab South applicable business model design and Asia, mLab East Asia, and (earlier) business analytics toolkits for mLabs mLab Southern Africa. In hindsight it and mHubs. In particular, the toolkits became clear that infoDev should have are envisaged to give advice on which helped incoming mLab managers and business models are most appropriate consortium representatives to under- under which ecosystem conditions and stand an mLab’s mission and infoDev’s within which innovation enabler land- goals much more swiftly. For future scape, as well as what roles different mLab grants, infoDev will consider stakeholder groups should ideally take setting aside additional staffing and (including the design of public-pri- 65 travel resources in order to follow up vate partnerships). These knowledge products are expected to help mLab and or policy constraints) continue to ex- mHub managers make better-informed ist. Importantly, there are indications decisions for their strategy and opera- that the tech innovation community’s tional models. They are seen as starting strong sense of identity combined with points to empower mLabs and mHubs skepticism towards government and to improve their monitoring and as- large corporations can have negative sessment of results and impact, in turn externalities, such as a lack of linkages helping them to learn about and show- to non-technical stakeholders or struc- case successes to funders and partners. tural underestimation of advocacy and Critically, such toolkits and analytical lobbying towards governments. products need to be highly customizable and interactive if underutilization is to be In line with this, improved benchmarking avoided, as was the case with the gener- of technology innovation hubs and other ic mLab business plan (see Evaluation, support institutions is needed to under- Business Analytics, and Toolkits). stand how roles in innovation ecosystem support can be divided up effectively. Areas for Future Research This report has only hinted at one of and Analysis many dimensions in which the strategies Finally, the report has brought to bear of innovation support institutions can several other areas that merit more differ (ecosystem builder versus startup in-depth research and analysis. These accelerator focus), but many other dis- topics could be developed by infoDev tinction factors determine whether the but also by other organizations working value propositions of organizations will to understand and support innovation complement and integrate with each ecosystems in developing countries. other (for instance, sector-specific in- novation hubs for mobile health, mobile With the recent surge of technology in- agriculture, and so on; hardware versus novation hubs across developing coun- software innovation hubs; ICT versus tries and in particular across Africa, an other technology innovation hubs; tech- increasing need to understand mecha- nology versus non-technology innova- nisms of collaboration, competition, and tion hubs; and so forth). differentiation between these hubs has emerged. This is particularly relevant Another important line of inquiry con- in vibrant ecosystems such as Nairobi, cerns the feasibility of standardizing Lagos, or Cape Town, where myriad, at least certain technology innovation mostly scattered support institutions hub functions across geographical are active. Interviewed stakeholders ex- boundaries. Evidence is mixed; while pressed a sense that these institutions some of the entrepreneurs’ demands ought to find better ways of coordina- concerning innovation ecosystem sup- tion, since demand for support services port appear to be the same across con- is high but there is also a significant texts, other constraints vary depend- risk of duplication and infighting while ing on local conditions.48 This raises structural and deep-running ecosystem the question of whether a franchise 66 barriers (such as basic skill shortages model for technology innovation hubs, in which hubs share a common brand concept. The report shows that, de- and the same basic functional model pending on the local context and stra- across a region or globally, could make tegic choices, mLabs and mHubs can their rollout and implementation more implement a variety of business mod- efficient. This report informs the de- els, and experimentation, adaptation, bate, but the evidence from mLab and and learning will have to continue. mHub pilots so far remains inconclu- Among the most pressing challenges sive for which elements of tech inno- for mLabs and mHubs is their uncer- vation hubs’ standardization is feasible tain financial sustainability. None of and for which it is not. the mLab pilots has found a business More broadly, the features and evolution model that could already cover core of app economies in developing coun- overhead costs as well as incubation tries still remain poorly understood. and acceleration services independent- For instance, there is great interest on ly from donors, impact investors, and the potential share of the app economy governments. This experience is simi- within the overall economy, and how lar to the overall experience of business it interacts and catalyzes other areas incubators in developing countries that of economic activity, in particular job are barely able to survive on service creation. Such research would inform revenues alone. mLab East Africa’s pri- the technology innovation hub agenda vate-sector sponsorship model has so insofar as it could further corroborate, far been the most promising approach or question, their relevance. in this direction but, without additional donor funding, the mLab might soon Conclusion have to eliminate any service line that The report concludes that the mLab does not generate a positive margin by and mHub pilots have established itself. Ecosystem builder services are proof for the viability of the mobile inherently hard to monetize through di- entrepreneurship enabler concept: rect contributions from clients and are mLabs and mHubs are capable of cre- likely to need donor and government ating and supporting growth-oriented contributions for extended periods of startups and filling gaps in innovation time. But also the setup of profitable ecosystems of developing countries. All “pure” startup incubators and acceler- mLabs and mHubs faced implementa- ators such as Raizcorp49 or Y-Combinator50 tion challenges and the experience has still seems unrealistic for the early shown that the rollout of such complex, stage innovation space of developing multistakeholder operations is never countries, where startup success rates, a given. Yet, by and large, most cases valuation potential, and profitability have fulfilled or exceeded expectations will always be lower than in more ad- (see part III for details). At the same vanced ecosystems and for higher-up time, mLabs, mHubs, and infoDev have innovation stages. It will be crucial for to continue to learn before they are mLabs that they continue to evaluate able to maximize the potential impact the potential of donor-based revenue 67 of the mobile entrepreneurship enabler models but also begin to experiment with success-sharing schemes, where competitions, more structured broker- supported startups either pay revenue age of partner and funder contacts, ap- or profit-based success fees, or give up plicable analytical products and bench- equity shares. marks, and active knowledge transfer in cases of local challenges are likely to Other important challenges were of- create a compelling value proposition ten due to administrative barriers and for mLabs and mHubs that are part of short implementation timelines, and infoDev’s Mobile Innovation network. these are areas for more straight- forward improvements. In particular, Of course, even if resources and as- future mLabs and mHubs will benefit sistance provided by infoDev were from more elaborate upfront scoping abundant, mLab managers would still assessments and high-level project have to be clear about their strategy design conducted by infoDev over one and address the many open questions to two years, before mLabs and mHubs of how to effectively impact innovation start implementation themselves. They ecosystems, some of which this report will ideally be endowed with sufficient has highlighted. infoDev, mLabs, and core funding over at least six years of mHubs will have to continuously evolve implementation, which should provide their practices and learn from past time to build up donor and govern- mistakes to improve new initiatives. The ment-independent revenues from the next months and years will tell whether private sector. the value proposition of infoDev, mLabs, and mHubs are in fact compelling for infoDev is also learning about mLabs’ their donors and partners, as well as and mHubs’ expectations towards its the startups, entrepreneurs, and devel- function as a convener and network opers that they serve. 2 68 broker. Selected global innovation Endnotes 23 As a case in point, mLab East Africa’s Pivot East competition has so far attracted the most significant contributions from the private sector, fueled by the competition’s high brand value and the rapidly growing East African innovation ecosystem. 24 http://www.mobilemonday.net/. 25 This also underscores that results cannot be compared directly across mLabs: in part due to selection of more seasoned entrepreneurs, mLab East Africa dwarfs mLab ECA with regard to investments raised, revenues generated, and jobs created by its startups (see the respective case studies for mLab East Africa and mLab ECA). 26 In this context, social enterprises are defined as startups that are profit-oriented, but ultimately see profit as a means to the end of rolling out an app or a mobile software product that has positive social impact. Typical sectors of interest include mobile health, agriculture, education, and government. 27 http://www.itwebafrica.com/ict-and-governance/523-africa/231471-africas-tech-hubs-are-they-producing-success-stories. 28 For a recent op-ed on the negative externalities of a narrow social development perspective on innovation and technology adaptation, see http://www.nytimes.com/2013/09/17/opinion/let-the-poor-have-fun.html. 29 Of course, the capital required varies from market to market and depends on the startups’ capital intensity. In the mobile app space of developing countries, typical preseed financing gaps start as low as $5,000 and reach up to $50,000. 30 Recently, significant advances have been made towards coherent methodologies and frameworks for the assessment of en- trepreneurial ecosystems that have helped infoDev to fine-tune its assessments. Notably, the World Economic Forum (2013) together with Stanford University and others suggested a framework of entrepreneurial ecosystem pillars and entrepreneurs’ perceptions of their relative importance. The study found that, globally, entrepreneurs perceive accessible markets, human capital and the local workforce, as well as funding and access to finance as the most important ecosystem pillars. Also GSMA’s Mobile for Development unit as part of its M4D Network program will publish a mapping of Kenya’s entrepreneurial ecosystem in early 2014, at https://mobiledevelopmentintelligence.com/. GSMA intends to carry out similar research in West Africa, South Africa, as well as South East Asia. 31 In 2013, CIDA was integrated with the Canadian Department for Foreign Affairs to form the new Department of Foreign Affairs, Trade and Development of Canada. 32 http://www.infodev.org/infodev-files/resource/InfodevDocuments_1087.pdf. 33 http://theleanstartup.com/principles. 34 The evaluation is expected to be released in March 2014, see http://www.infodev.org/mobile. 35 Country case studies were created for Kenya (http://www.infodev.org/articles/mobile-usage-base-pyramid-kenya) and South Africa (http://www.infodev.org/articles/mobile-usage-base-pyramid-south-africa), followed by an overview report (http://www.infodev.org/highlights/mobile-usage-base-pyramid). 36 http://www.globalforum2013.co.za/. 37 http://www.infodev.org/articles/open-innovation-africa-summit-2. 38 http://www.infodev.org/m2Work. 39 http://www.infodev.org/mAgri. 40 http://ventureoutchallenge.org/. 41 http://www.infodev.org/articles/mobile-usage-base-pyramid-kenya. 42 http://www.infodev.org/articles/mobile-usage-base-pyramid-south-africa. 43 http://www.infodev.org/highlights/mobile-usage-base-pyramid. 44 http://www.infodev.org/articles/tor-virtual-incubation-pilot-vietnam. 45 http://www.infodev.org/articles/east-african-virtual-incubation-pilot-launch-nairobi-kenya. 46 http://www.appcampus.fi/. 47 A first step in this direction has been taken by GSMA’s Mobile for Development unit. As part of its M4D Network program, GSMA interviewed about 300 organizations, analyzing the state and future of the entrepreneurial ecosystem in Kenya, including benchmarking of prominent tech innovation hubs. Findings were published in early 2014 at http://www.gsmaentrepreneurshipkenya.com/. 48 See, for instance, World Economic Forum [2013). 49 http://www.raizcorp.co.za/. 50 http://ycombinator.com/. 2 69 Part III: Case Studies of mLab and mHub Business Models Applying the Business Model paid the mLab or mHub to implement Canvas to mLabs and mHubs a specific activity. For these clients, the mHub or mLab can be seen as a supplier T he following case studies, as the or service provider. Sponsors and do- final part of the report, employ nors (including private-sector sponsors, the Business Model Canvas51 (see contributing government agencies, and figure 10) to analyze and illustrate four infoDev) were included as partners but mLabs and three mHubs in depth. This not as customer segments, as these chosen because of its illustrative power stakeholders do not request a specific and its popularity among startup com- service or deliverable in return for the munities. General guidance on how to funding they provide.54 apply the Business Model Canvas can In order to generate valuable lessons be found at various sources online52 that are lost in the Business Model or in the original book by Alexander Canvas perspective, each of the seven Osterwalder.53 The planned but never in-depth case studies also discuss the implemented mLab South Asia, as well challenges that the mLab or mHub as Mobile Monday Kampala (mHub in question has faced. Stakeholders Uganda) are included as essays with- also requested that results should be out discussion of the Business Model displayed and benchmarked, even if Canvas framework; these cases held simple quantitative measurements valuable insights but data collection have limitations (see Why Evaluate that would have been necessary for the Business Models of mLabs and more in-depth analysis was not feasible. mHubs? and Results for mLabs and It should be noted that the Business mHubs), and also thought that mLabs’ Model Canvas has major limitations. In and mHubs’ financial sustainability and the context of mLabs and mHubs, the their potential to become independent most important conceptual problem of government and donor support were concerns the categorization of stake- important areas of interest—each in- holder segments that benefit from, but depth case study also includes sections also contribute to, the value proposition; on these topics. The results sections for these stakeholders can be categorized as the four operational mLabs also pres- both partners and customer segments ent radar charts that give an indication (and sometimes also as resources). how the given mLab’s results compare The following case studies consider the to other network members (see Results developers, entrepreneurs, and startups for mLabs and mHubs for further com- that mLabs and mHubs serve as custom- parative analysis and appendix C for 2 70 er segments, and also organizations that details on the calculation). KEY PARTNERS KEY ACTIVITIES VALUE PROPOSITIONS CUSTOMER CUSTOMER SEGMENTS RELATIONSHIPS - Who are our Key Partners? - What Key Activities do our Value - What value do we deliver to - What type of relationship does each of For whom are we creating value? Propositions require? the customer? our Customer Segments expect us to Who are our most important - Who are our key suppliers? establish and maintain with them? customers? - Our Distribution Channels? - Which one of our customer’s - Which Key Resources are we problems are we helping - Which ones have we established? acquairing from partners? - Customer Relationships? EXAMPLES: to solve? - Which Key Activities do - Revenue streams? - How are they integrated with • Mass Market - What bundles of products and the rest of our business model? • Niche Market partners perform? CATERGORIES services are we offering to each • Segmented MOTIVATIONS FOR PARTNERSHIPS Customer Segment? - How costly are they? • Production • Diversified • Optimization and economy • Problem Solving - Which customer needs are EXAMPLES • Multi-sided Platform • Reduction of risk • Platform/Network we satisfying? and uncertainty • Personal • Automated CHARACTERISTICS assistance Services • Acquisition of particular resources and activities • Newness • Dedicated • Communities • Performance Personal • Co-creation • Customization • Self-Service • Getting the Job Done • Design • Brand/Status KEY RECOURCES CHANNELS • Price • Risk Reduction • Accessibility - What Key Resources do our Value Through which Channels do our Customer The Business Model Canvas by Alexander Osterwalder • Convenience/Usability Propositions require Segments want to be reached? - Our Distribution Channels? How are we reaching them now? - Customer Relationships? How are our Channels integrated? - Revenue Streams? Which ones work best? TYPES OF RESOURCES Which ones are most cost-efficient • Physical How are we integrating them with • Intellectual (brand patents, customer routines? copyrights, data) CHANNEL PHASES • Human 1. Awareness 4. Delivery • Financial 2. Evaluation 4. After Sales 3. Purchase COST STRUCTURE REVENUE STREAMS - What are the most important costs inherent in our business model? - For what value are our customers TYPE FIXED PRICING DYNAMIC PRICING really willing to pay? • Asset sale • List Price • Negotiation - Which Key Resources are most expensive? - For what do they currently pay? • Usage fee • Product feature (bargainning) - Which Key Activities are most expensive? • Subscription Fees dependent • Yield - How are they currently paying? • Lending/Renting/ • Customer segment Management TYPES OF RESOURCES SAMPLE CHARACTERISTICS - How would they prefer to pay? Leasing dependent • Real-time- • Cost Driven (leanest cost structure, low price value • Fixed Costs (salaries, rents, utilities) • Licensing • Volume dependent market proposition, maximum automation, extensive outsourcing) • Variable costs - How much does each Revenue • Brokerage fees • Value Driven (focused on value creation, premium • Economies of scale Stream contribute to overall • Advertising value proposition) • Economies of scope revenues? Source: http://upload.wikimedia.org/wikipedia/commons/1/10/Business_Model_Canvas.png 71 2 Figure 10: Osterwalder by Alexander The Business Model Canvas 72 KEY PARTNERS KEY ACTIVITIES VALUE PROPOSITIONS CUSTOMER CUSTOMER SEGMENTS RELATIONSHIPS • Network & brand partner (iHub) • Promotion, introductions, & • Connections & opportunities • Personal, direct assistance for • Incubatee startups exposure for startups (partners, customers, networking) business matters & networking & entpreneurs • Improved business model • Startup teams • Incubation & training partners • Niche, self selection of startup teams (Viktoria Solutions, eMobilis, Web through mentoring & coaching with minimum viable app product that Foundation, AfriLabs, Nokia) • One-on-one mentoring & • Subsidized office space & seek investment & exposure coaching, workshops equipment (connectivity, • Student tech community • In-depth technical & app testing) • Niche, self selection of technical • Tech partners (Intel, Microsoft, entrepreneurship training Qualcomm, Samsung, Nokia, & business students in need for • Training & developer outreach InMobi, MIH) skills to start an app company • Pitching & business skills programs (Intel, Microsoft, • Branding and outreach, Samsung, Nokia, Qualcomm contact brokerage • Networking opportunities • Personal & brokered, some • Ad hoc partnerships for & contacts communication is managed by iHub competitions, projects, events, & • Competitive pressure, exposure outreach (Uganda Communications Commission, Outbox Uganda, KEY RECOURCES CHANNELS CGAP, Blackberry, MTN, Samsung, The Business Model Canvas – mLab East Africa USAID, GrowVC, Young Innovations, • Technical & entrepreneurial infoDev, University of Nairobi) • Office & meeting spaces, connectivity skills & knowledge • Personal interaction with in Bishop Magua Building management & mentors • Locally recognized certificate • iHub’s brand, outreach channels, • Outreach around Pivot East through networks, & resources (expertise, staff) social media, iHub blog, mailings • Access to developer community, experience in ecosystem • Personalized communication & outreach & training personal interaction at mock pitch • Entrepreneurship & technical sessions for Pivot East Finalists expertise, app testing facility • Personal interaction with • Established but evolving & eMobilis trainers flexible training curriculum, • University outreach & social experienced teachers media for trainee recruiting • Sponsorships, in-kind contributions, • Personal interaction & iHub media contacts, networks, & brands & staff (social media, informal meetings, briefings, negotiations) COST STRUCTURE REVENUE STREAMS • Annual partnerships (Mocality International Holdings (MIH), inMobi, • Coaching & Advisory • Seed funding • Incubation revenue Nokia, Samsung, Microsoft) • mLab eMobilis training • Fixed cost (lease, staff, etc.) • Events (sponsorships, participation fees) • Trainings, service partnerships • infoDev grant, tranche from virtual incuba- • Events, ecosystem building tion grant (administered by AfriLabs) Note: To access the full canvas for mLab East Africa, go to https://bmfiddle.com/f/#/rW926. Each and relationships, channels, and value proposition elements have the according color when they partner or group of partners has a specific color; activities and resources that mainly rely on a mainly relate to the customer segment in question. Activities and resources as well as cost and particular partner group have the same color. Also each customer segment has a distinct color, revenue streams that only relate to the infoDev grant have their own color. A flexible, busy incubator for a bubbling startup ecosystem mLab East Africa Total infoDev grant funding $725,000 Funding for technical assistance $194,000 (approximate) Base city, country Nairobi, Kenya infoDev grant start date November 1, 2010 First activities launched March 1, 2011 Official launch June 16, 2011 Consortium members iHub (lead) eMobilis (coleader) World Wide Web Foundation University of Nairobi Key partners Intel, Microsoft, Samsung, Nokia, Qualcomm, InMobi GrowVC, USAID Targets as per grant agreement • Reaching a minimum of 100 developers through outreach and capacity building workshops (bootcamps), held in Uganda, Tanzania, and Rwanda • 150 students trained in mobile app development, business and entrepreneurship • Fifteen trained incubatees with products brought to market, including ten startups hosted within the mLab Website http://www.mlab.co.ke/ Social media and other web resources @mLabEastAfrica, http://www.pivoteast.com/, Table 3: Basic @PivotEast, https://facebook.com/mlabeastafrica, Data for mLab http://www.youtube.com/user/mlabeastafrica East Africa Note: Funding amounts are based on actual expenses as of September 2013. Amounts are rounded to thousands. Business Model also started complementary com- mLab East Africa is the longest run- munity building, such as the Wireless ning and probably the busiest of all Wednesday series. mLabs. The mLab is located in the Bishop Magua Building in Nairobi, one The incubation program benefits eight floor below the now-famous iHub.55 resident startups, with offices in the The iHub community and the vibrancy Bishop Magua Building, for up to two of the local innovation ecosystem have years. The interviews, in line with re- attracted international attention and cent findings of an evaluation by the the mLab’s services have been in high University of Nairobi,58 showed that Figure 11: demand. Its core activities include the way in which the mLab adds value (opposite page) an incubation program, the annual could differ greatly from startup to Business Model regional startup competition Pivot startup. Most startups mentioned one- Canvas for mLab East,56 and an intensive four-month on-one business mentoring and coach- East Africa 73 training program. 57 Later, the mLab ing as the most important service—in iHub’s Influence, from mLab East Africa’s perspective The iHub has become one of the most widely known innovation hubs in Africa. In recent years, not only has the number of iHub community members grown far beyond 10,000, the innovation space has also attracted interest from many high-level tech executives and thought leaders from across the globe. Companies and organizations that look for channels to engage with developers and the tech community in Kenya often see the iHub as a powerful one-stop shop. The iHub’s mission is to strengthen the ICT ecosystem of Kenya and East Africa. As the consortium leader, the iHub sees mLab East Africa’s niche focus on startup creation as one important component among its several other work streams. The iHub’s clout has undoubtedly been instrumental in increasing the mLab’s reach, securing contributions from large tech partners (for instance, Intel, Microsoft, Samsung, Nokia, and Qualcomm), and engaging a greater number and variety of stakeholders in mLab projects and activities. Occasionally, the iHub and its subsidiaries have also assisted the mLab’s startups directly with contact brokerage, accounting support, or workshops on market research. How mLab’s affiliation and colocation with the iHub materialized as added Box 5: value for individual startups depended on their needs and priorities. Startups iHub’s such as Kopo Kopo, MedAfrica, mFarm, Eneza Education (formerly mPrep), Influence, from mLab or Zege Technologies—which rely on widespread awareness for their product, East Africa’s including among social development organizations and impact investors— perspective have probably benefitted the most. particular, the advice from Viktoria Also the value of the infrastructure that Solutions, a local consultancy that the the mLab provides (office space, con- mLab hired.59 nectivity, app testing facility, meeting rooms, some overhead expenses) is Another key element of the value prop- important to most incubatees. Annual osition consisted of networking oppor- funding or device donations from tech tunities and exposure. At times, mLab partners like Intel, Microsoft, Nokia, and iHub staff would make introductions Samsung, Qualcomm (and earlier on based on the startups’ needs. But often InMobi and MIH) help cover or limit the it was simply brand affiliation with the mLab’s expenses, inter alia, in exchange mLab, the iHub, and the World Bank, for prominent branding in the mLab’s in combination with the iHub’s effective offices. The mLab’s agreement with marketing, which enabled the startups to the iHub to share resources such as be heard by larger institutional partners accounting has also helped several en- and then pursue agreements by them- trepreneurs to improve their core busi- selves (see box 5). Being in the midst of ness functions. Again, the significance the iHub community also made it easier of the value varies between startups and for some startups to source talent and entrepreneurs. For instance, app testing 74 serendipitously learn from peers. is crucial for trainees and entrepreneurs By putting startups under competitive Another critical value proposition that pressure while Pivot East offers consists of networking supporting them and partnership building, at least for the 25 finalists. Several interviewees at the same time, mentioned that they found partners or Pivot East has increased their network. The goodwill proven to be a good and brand value that Pivot East was able to establish fairly quickly enabled incubatee selection organizers to gain large sponsor- channel. ships and contributions from partners such as Intel, Samsung, Microsoft, Nokia, Qualcomm, the Ugandan Communications Commission, Outbox with new apps, but it might only be mar- Hub Uganda, and others. Also mLab ginally relevant for those startups that consortium members iHub and eMobi- have already rolled out a well-function- lis have supported Pivot East, taking on ing app to multiple platforms. marketing and event organization tasks. Pivot East, as mLab East Africa’s sec- The third important pillar of mLab East ond activity pillar, has provided value Africa’s value proposition is skills de- beyond seed funding of five category velopment and startup creation through winner startups per year. One effect has an intensive training program, mainly been startup creation in high numbers: targeted at graduates from technical or all applicant teams have to formalize business-oriented courses at Nairobi’s their startups. Naturally, a formalized universities. The training program is company is not necessarily a sustain- implemented by consortium coleader able company, and so 25 finalists re- eMobilis, while some modules were ceive additional coaching, in particular, delivered by Viktoria Solutions (on en- on pitching. Interviewees pointed out trepreneurship) and the Web Foundation that the feedback received—even if it (on user experience and user interface can feel harsh to young entrepreneurs design). The training initially had a at first—helps to improve not only the strong focus on technical skills; the goal pitch deck for the event, but also to was to complement university programs hone their proposed startup business with applicable, mobile software coding models. More and more, applicants skills. As organizers realized that, with have realized that they need to learn this setup, they could hardly spur on and improve through iteration, which trainees to found startups, they recruit- has resulted in a decreasing role for ed more students from business and prize money as a motivation for par- non-technical backgrounds, shifted fo- ticipation over the years. By putting cus towards entrepreneurship training, startups under competitive pressure and required trainees to engage in part- while supporting them at the same nership and business building for their time, Pivot East has also proven to be a final course project. Later, the trainee viable incubatee selection channel. teams’ quality improved to a level where 75 Overall Revenue generated reach by startups 100% 80% 60% Investments Events raised by startups 40% 20% 0% Applications Jobs created for support & supported Training Startups & startup teams created In depth App customer team support traction Figure 12: App innovation, (apps Comparative monetized, brought to Results for mLab market, prototyped) East Africa Note: The percentage values represent linear comparisons across data for all four mLabs, with the mLab scoring highest for a given indicator receiving 100%. Indicator Result as of June 2013 Revenues generated by startups $627,000 External investments raised by startups $1,500,000 Direct jobs created by startups 100 Startups created 50 Solid startup teams that have not registered a business 50 Consumer customer traction: Number of app downloads/users/subscribers 263,000 Organizational customer traction: Number of enterprise and public agency customers 5447 Apps brought to market 196 Number of other app prototypes 301 Number of teams applying for in-depth support 460 Number of individual applicants to services 900 Teams that received in-depth support 119 Table 4: Results Number of people trained 462 Highlights for Number of events 55 mLab East Number of developers and entrepreneurs reached 5,000 Africa 76 Note: As reported by mLab manager. Including estimates, proxies, and underreporting due to missing data. some of them qualified as Pivot East and employed about 100 people. The finalists. Throughout, tech partners in- amounts of money that has been cluding Samsung, Microsoft, Qualcomm, invested in startups is not evenly dis- and others contributed to the training, or tributed; startups such as Kopo Kopo ran their own platform-specific training raised hundreds of thousands of dol- in collaboration and coordination with lars, while some of the teams collect- the mLab. ed seed money sums as low as a few thousand dollars, often in the form of The fourth and final activity pillar grants. For instance, four teams re- emerged when the mLab found that ceived impact funding of about $20,000 the iHub tech community was lacking following the Wireless Wednesday sub-communities for specific subject series on mobile agriculture that areas and sectors that include poten- the mLab hosted in partnership with tial startup customers and non-tech USAID. The mLab helped broker part- organizations. It set up the Wireless nerships with Samsung for mFarm Wednesday event series for mobile and Zege Technologies and with Nokia agriculture (sponsored by USAID), for for Whive. The tremendous demand mobile health (sponsored by Samsung), for the mLab’s services is illustrated mobile finance (in partnership with CGAP), tech education, and other by the high number of applications for topics. The mLab also co-organized support: 460 teams applied for in-depth Mobile Mondays and other communi- support (most for Pivot East) and about ty-building events as it saw fit. 900 individuals applied for participation in training, workshops, and events. Results Challenges Within a little over two years of opera- tion, mLab East Africa has graduated Many of mLab East Africa’s challenges ten resident startups, hosted three stem from insufficient capacity and re- editions of Pivot East as a major startup sources to guarantee depth of service competition for the region, trained five delivery and follow-through on the many waves of trainees, launched several opportunities available in the vibrant small series of Wireless Wednesday Kenyan mobile innovation ecosystem. community events, and run many other In particular, the mLab manager was ad hoc initiatives such as workshops, expected to assume an overly broad hackathons, and gatherings. As with array of roles and responsibilities, other mLabs’, mLab East Africa’s results ranging across fundraising, strategy, need to be seen against the backdrop of operations, mentoring, contact bro- the local innovation ecosystem; Nairobi kerage, and relationship management. has a buzzing scene but there is still only Several interviewees involved in the a small number of mobile app startups mLab’s management and governance that have received venture capital.60 thought that the size of the grant should allow for hiring two managers, with In total, the mLab’s incubatees have one representing the mLab externally made about $627,000 in revenue, (building partnerships, fundraising) and 77 raised $1.5 million in investments, one administrating the mLab’s affairs. Noted startups and app products supported by mLab East Africa Box 6: Noted start -ups and app products supported by mLab East Africa Note: Websites at http://www.kopokopo.com; http://mfarm.co.ke; http://enezaeducation.com; http://zegetech.com/ portal/. A more comprehensive list of supported startups can be found in the appendix. Detailed case studies of these and other startups will be featured in an upcoming infoDev publication, commissioned to the private sector development evaluation and research agency CAD (http://cad.globalcad.org/en/que-es-cad/). These interviewees thought that this Kenya, although efforts to strengthen setup would also help to address crit- the mLab’s regional virtual incubation icism from outsiders that the mLab’s program for Pivot East finalists (see current leadership was falling short Sustainability and Next Steps) could on fundraising and external represen- change this diagnosis. tation and that it did not sufficiently instill a business mindset for startups: Less intense support and misaligned in the past, a lack of resources had incentives at times resulted in low kept the mLab management from tight commitment and identification among monitoring and critical assessments of the mLab’s clients. For example, in- some of the startups’ business models cubatee startups were often not eager and monetization strategies. to participate in group workshops and peer-learning sessions. From their per- In line with this, interview participants spective, sparse one-on-one support highlighted that the mLab’s most im- limits the mLab’s value proposition. portant shortcoming in service delivery Accordingly, some interviewed startups was insufficient mentoring and coach- felt that the gain that they receive from ing. The mLab found it cost-prohibitive the mLab’s incubation services would to consistently engage qualified men- not justify success-sharing beyond the tors that could enhance the value of current rent payments. Of course, this incubation for startups. Workshops that assessment is made at a time where aimed at coaching several startups at the startups have already received the same time were considered useful support for free, which, as other in- only in exceptional cases. The lack of terviewees pointed out, might have led mentors was also seen as one reason to a sense of entitlement and lowered 78 for the mLab’s lagging footprint beyond appreciation for the intangible assets that the startups built through the Pivot with niche business models were not East competition and other services able to benefit much from serendipitous received. Also trainees lamented that networking enabled through the iHub participants should be selected more as much as might have been expected. purposefully for them to be able to build They rather experienced distraction by viable startup teams, and also the lack an influx of visitors who would seldom of one-on-one follow-up after grad- offer concrete business opportunities uation from the training was seen as for them. Interviewed iHub outsiders limiting—although these interviewees and some evidence from the University also acknowledged that improvements of Nairobi’s evaluation of the mLab would require a costly additional effort pointed to the more general observation on the mLab’s part. The mLab saw train- that several mLab teams and startups ees drop out of the free classes during are disengaged from market needs, as the first training waves in 2011 and they overly focus on their technology 2012, leading it to introduce a fee that (and not their business model), become would be partly reimbursed only upon hyped too quickly, or fail to venture out submission of final course projects. of the iHub “bubble” to experience and understand their customers’ problems. In turn, the mLab and some of the entre- Owing to a lack of capacity, the mLab preneurs believe that seed investments was also not able to tap into infoDev’s would be a good complement to the network and work closely with other mLab’s current services, and that this mLabs and mHubs. would also increase the willingness to share equity or pay royalties. Yet, such Similarly, the mLab felt that it lacked the investments should be coupled with resources to maintain and improve its ongoing support, for instance, through value proposition for paying tech part- regular monitoring and step-wise fund- ners (such as annual mLab and Pivot ing based on milestones. The Pivot East East sponsors) over time. The mLab editions of 2011 and 2012 had shown experienced that, once the branding ef- that any direct financial rewards should fect of a partnership announcement had always have strings attached and be been achieved, some sponsors would complemented by mentoring, even if it ask for more concrete deliverables implies more necessary follow-up for and at least a certain degree of brand the mLab: without follow-up, several exclusivity, or they would lower their Pivot awardees had spent the prize mon- contribution for ensuing sponsorships. ey on things other than their business. Sponsors’ stretched resources aggra- vated the situation: developer outreach More broadly, interviewees suggested units are not often endowed with many that the mLab improve the targeting personnel or significant budgets. and rigor of network building for clients. Interviewees pointed out that the mLab Moreover, for most tech sponsors, de- could set up structured collaborations veloper outreach consists of marketing with organizations that have mobile and technical training rather than start- market intelligence, or other innovation up incubation, and it was sometimes 79 hubs in Nairobi. In particular, startups unclear to them how the mLab can add 2013 2014 2015 Expenses Events and ecosystem building 132,500 139,125 146,081 Seed funding (investments) 47,500 125,000 125,000 Lease, personnel, training, coaching and advisory, other 299,000 293,200 310,020 Total expenses 479,000 557,325 581,101 Income Services and sponsorships Annual partnerships 60,000 66,000 72,600 Incubation revenues 34,000 24,000 28,800 Events, ecosystem building, training, other 210,000 221,500 233,675 Grants Table 5: infoDev 100,000 — — Budget and Financial Total income 404,000 346,500 370,075 Projections for mLab Financing gap (75,000) (210,825) (211,026) East Africa Total financing gap until 2015 (496,851) Note: All amounts in US$. Budget items are kept at a high level to avoid disclosure of sensitive information. Projections are based on rough estimates, as mLabs are currently in a transition phase at the end of grant funding from infoDev. value beyond what the iHub can provide. supervision function as consortium or- In turn, other interviewees claimed that ganizations. This led to conflicts of in- the training program should be better terest, for instance, when project con- integrated with the mLab’s overall tracts were tendered. The consortium goals, the incubation program, and was mostly able to mitigate any severe Pivot East; the startups that trainees conflicts by maintaining relationships founded were often not sustainable, of mutual trust and support among its only few trainee teams qualified as Pivot members, but interviewees expressed East finalists, and incubatee startups the view that this was only possible only rarely sourced talent from among because the consortium organizations trainees. Interviewees pointed out that and the mLab itself were at least in the training program might well have a part driven by an ecosystem-oriented substantial positive impact for the eco- mission, and not profit. Some consor- system and startups in the long run, but tium members mentioned the lack of the lack of immediate results and the direct compensation for their contribu- fact that training is hardly generating tions—especially in the mLab’s setup revenue made it difficult to justify main- phase—as another challenge. taining this activity. Finally, some client entrepreneurs and Another challenge consisted of subtle tech partners pointed out that infoDev’s difficulties with the mLab’s consor- role was not clear to them, and that tium structure. At different points in they would like to know more about time, consortium members effectively other mLabs and mHubs. These stake- became service providers for the holders also looked to infoDev to build 80 mLab, while they still assumed their partnerships with global organizations and bring mLab entrepreneurs togeth- strategy over the long run. As a result, er for peer-learning and international the mLab is still looking to make up for exposure (for instance, through boot- projected funding gaps over the next camps with participation from across few years (see Table 5). the network). On the other hand, both the consortium leader and the mLab An equity investment program for the management pointed out that infoDev’s incubated startups that is currently in hands-off approach towards local the design phase is likely to contribute operations and time pressure initially income, albeit equity liquidations are enabled a quick and effective rollout of expected to generate returns only after activities; avoiding government or do- several years. This goes hand in hand nor intervention was an explicit focus in with a redefinition of the mLab’s core their grassroots-driven approach. Like value proposition: it aims to move away other mLabs, mLab East Africa felt that from a focus on physical space and reporting to infoDev and adherence to colocation with the iHub as core assets World Bank grant and procurement towards emphasizing intangibles that rules was arduous, both for the mLab are generated through mentoring and and for entrepreneurs. network brokering. This reorientation is also a reaction to incubatees’ opinion Sustainability and Next Steps that the value that the mLab currently adds to their business would not justify mLab East Africa has opted for a pri- an equity dilution and that the existing vate sector-oriented approach based rent model is more appropriate. on technology partnerships and spon- sorship, without seeking government The virtual incubation program that funding. In particular, the mLab was the mLab is currently piloting in col- able to attract large contributions for laboration with AfriLabs61 is envisaged Pivot East, from annual memberships of to bolster the mLab’s strategic shift, tech partners, as well as joint execution and it should also contribute towards of developer training. The mLab was the mLab’s goal to become a mobile in- able to significantly diversify its revenue cubator for the whole East African re- streams; for 2013, the mLab projected it gion. In the same context, the mLab is would generate roughly three-quarters looking to synchronize the annual cycle of its revenue from sources other than of incubatee graduation and selection the infoDev grant. However, the mLab’s with the Pivot East conference, so that training program is an example of an at least most of the available virtual activity that is widely seen as crucial to incubation slots can be filled with Pivot infuse further applied entrepreneurship East finalists. Another change will and technical skills into the ecosystem, be further adaptation of the training but which is unlikely to generate profit. program into a module-based mini-ac- Moreover, some sponsors have become celeration program, in order to both more demanding over time and often satisfy calls for more entrepreneurship do not support mLab’s overhead and in- focus and to align the training program cubation costs, so that strong reliance better with other activity pillars. 81 on sponsorship could be a precarious 82 KEY PARTNERS KEY ACTIVITIES VALUE PROPOSITIONS CUSTOMER CUSTOMER SEGMENTS RELATIONSHIPS • Consortium members • Structured incubation program (Ideas • Brokerage of connections & • Incubated startups (focus on CSIR Meraka Institute, Lab, Acceleration Lab, Launch Lab) opportunities (partners, customers, • Personal, direct assistance for The Innovation Hub) financing, exchange with other business matters & networking incubatee startups, innovation • Student tech community • Space & connectivity, admin & competitions, networking) • Department of Science & procurement, IP & technical expertise, • Niche, self selection for • Mentoring & coaching Technology (DST) of South Africa, outreach, exchange on opportunities interesting events • Training & developer outreach Gauteng Province for clients • Lowered overhead (office space, programs (Nokia, Vodacom) connectivity, app testing, marketing, admin) • Personal, mLab management • App development clients • Ad hoc partnerships for events, • Integration & coordination with sets up contracts, brokers competitions, & outreach (Gauteng national & provincial ICT strategies relation to trainers Province, Department of Science • Competitive exposure & networking • Gauteng Province (managed and Technology, Nokia, Microsoft, (innovation competitions through TIH) Qualcomm, BlackBerry, Dimension • Outreach, innovation competitions, & hackathons) • Brokered, The Innovation Hub Data, Balefyre, Vodacom, Appcampus, events, projects (Gauteng Innovation manages relation for mLab open. NASA, infoDev, eCITI (Eastern Competition, m2Work hackathon, Urban • Awareness, basic Cape), Geekulcha, Appchemy, App Challenge, MTN App of the Year community building mFactory, Innoprenez, Jatamobile, Award, NASA Space Apps Challenge, Ungana Afrika, SAIS, STIFIMO, SDK workshops, Nokia & Vodacom CHANNELS MTN, SAINe, University of Pretoria, trainings, mLab Studio, Microsoft • Access to developer The Business Model Canvas – mLab Southern Africa Tshwane University of Technology, Student Partners, Open Doors at mLab) community, experience Durban University of Technology, • Personal interaction with University of Capetown, Technology management & mentors Innovation Agency) • “Vertical model”, mLab adds mobile • Local economic & skills tech expertise to existing incubatoin development, thought leadership & co-creation activities • Calls for participation in • Tech hubs & incubators in South competitions & events through Africa (Bandwidth Barn, Raizcorp) social media, mailings, KEY RECOURCES university outreach • Manager’s wide network in mobile • Sponsorships & project funding, • Personal interaction with tech, training, incubation technical expertise, subject-matter managers & trainers expertise & exposure, access to developer sub-communities, etc. • Experience with tech incubation projects • Regular updates through The • Established incubation programs Innovation Hub leadership (incl. business mentors & coaches, • Funding, customer leads for incubatees infrastructure, etc.) COST STRUCTURE REVENUE STREAMS • Developer ecosystem outreach • Travel • Provincial & national government • App development & consulting • Resident coach • Developer ecosystem outreach • Sundries • infoDev grant • Trainings (Nokia) • Office expenses & salaries Note: To access the full canvas for mLab Southern Africa, go to https://bmfiddle.com/f/#/Hd8G9. and relationships, channels, and value proposition elements have the according color when they mainly Each partner or group of partners has a specific color; activities and resources that mainly rely on a relate to the customer segment in question. Activities and resources as well as cost and revenue particular partner group have the same color. Also each customer segment has a distinct color, streams that only relate to the infoDev grant have their own color. A networked incubator, guiding mobile app startups from idea to launch at multiple locations mLab Southern Africa Total infoDev grant funding $380,000 Funding for technical assistance $194,000 (approximate) Base city, country Pretoria, South Africa infoDev grant start date July 31, 2011 First activities launched August 29, 2011 Official launch September 15, 2011 Consortium members CSIR Meraka Institute The Innovation Hub Ungana Afrika The Innovation Lab (lapsed) Key partners Gauteng Province, Department of Science and Technology of South Africa, Nokia, Microsoft, Qualcomm, BlackBerry, Dimension Data, Vodacom, Appcampus, CITI (Bandwidth Barn) Targets as per grant agreement • 50 students trained in mobile app development • Eight to ten incubatees with products brought to market • Five startups hosted within the mLab Website www.mlab.co.za Table 6: Social media and other web resources @mLabSA, Basic Data for http://www.youtube.com/user/mLabSAStudio mLab Southern Africa. Note: Funding amounts are based on actual expenses as of September 2013. Amounts have been rounded to thousands. Business Model to deliver training programs, for which mLab Southern Africa has become the mLab recruits highly skilled entre- a focused, highly networked, and preneurs from incubated startups as well-funded incubator and accelerator trainers to ultimately service student for mobile app companies. The mLab, technologists. even more than its peers, functions as a platform that interconnects client Accordingly, most interviewed startups groups in a multisided market. In other pointed out that brokerage of contacts words, most of the mLab’s customers and opportunities is the core value are also partners that it works with to that the mLab Southern Africa gener- service other customer segments. The ates for them. For example, the mLab core value proposition is targeted at has connected client entrepreneurs Figure 13: to external coaches and expertise (opposite page) the incubated startups; the startups’ Business growth in turn feeds into the value (for instance, through tech partners Model Canvas propositions for other client groups. such as Nokia) as well as financing for mLab For instance, developer outreach pro- and exposure opportunities (for in- Southern Africa grams of large tech companies like stance, through The Innovation Hub’s62 83 Vodacom or Nokia contract the mLab Gauteng Innovation Competition63). The mLab also introduced organizational staged incubation program: in the customers in search of a mobile solu- two month-long Ideas Lab, entrepre- tion to the right incubatee, and start- neurs brainstorm business models ups confirmed that the mLab’s brand or join hackathons, seeking feedback recognition helps them with customer from coaches and other startups in acquisition—so much so that some or- an open-ended process. Selected ganizational clients (in particular, large entrepreneurs then spend about four firms and the public sector) prefer months in the Acceleration Lab, during engaging directly with the mLab, which which the focus shifts to product devel- facilitates the project for the startups opment while distractions are reduced in exchange for a management fee. and exchanges with outside partners The entrepreneurs see this help as a are kept to the most critical ones. viable interim solution until they es- Lastly, startups go to the Launch Lab tablish their reputation independently. phase, where they bring their product The current mLab manager’s personal to the market and reach out to external network, established over years of accelerators, partners, and investors. work in the mobile app sector, was Throughout the process, a resident also described as a critical asset, and coach gives technical and business stakeholders appreciate his readiness development advice on a daily basis. to leverage his past contacts and make The focus is on project design, mon- introductions as appropriate. etization strategies, and a startup’s financial sustainability. The mLab Beyond opening doors for startups, manager has monthly sessions with the mLab also contributes mentor- the startups, discussing milestones 84 ing and coaching during a focused, and higher-level strategic decisions. This helps startups to create a sense (DST).65 DST’s financial support was of accountability while the manager re- seen as matching funding for infoDev’s mains aware of the startups’ progress grant. Hence the ministry is hands-off and needs. in terms of regular exchanges with the mLab: it explicitly does not seek to in- Furthermore, the mLab provides terfere with operations or set specific startups with core support features outcome targets.66 The Innovation Hub of an incubator, such as office space, as a subsidiary of the Gauteng Growth lower overhead cost and back office and Development Agency helped the support, connectivity, and app testing mLab (and indirectly some of its start- devices. The mLab’s location within ups) to maneuver the procurement The Innovation Hub science park has channels of the provincial government mostly been a challenge (see section as a client. The financial support that Challenges) but it has been advan- the mLab receives from the Gauteng tageous for those startups seeking Province is more akin to a contractual to work with government agen- agreement, with The Innovation Hub cies or mLab consortium member as an intermediary: the mLab received organizations. specific deliverables with a focus on skill and enterprise development. The Innovation Hub and the CSIR Through the partnership, the province Meraka Institute64 as the lead consor- aims to solidify its positioning as a na- tium members have also played a crit- tional leader in private-sector support ical role in managing the mLab’s re- for technology innovation and ICT- lationships with its two major funders enabled public service delivery. aside from infoDev. CSIR assisted the mLab with securing and administrat- As a unique case in infoDev’s mobile ing the contribution from South Africa’s innovation network, mLab Southern Department of Science and Technology Africa has also made an effort to Noted startups and app products supported by mLab Southern Africa Box 7: Noted startups and app products supported by mLab Southern Africa Note: Websites at http://www.gometro.co.za/; http://www.afroes.com/; http://www.aftarobot.com/; http://geekulcha. com/. A more comprehensive list of supported startups and teams can be found in the appendix. Detailed case studies of these and other startups will be featured in an upcoming infoDev publication, carried out by the private sector development evaluation and research agency CAD (http://cad.globalcad.org/en/que-es-cad/). 85 Indicator Result as of June 2013 Revenues generated by startups $345,000 External investments raised by startups $652,205 Direct jobs created by startups 51 Startups created 12 Solid startup teams that have not registered a business 21 Consumer customer traction: Number of app downloads/users/subscribers 598,000 Organizational customer traction: Number of enterprise and public agency customers 30 Apps brought to market 115 Number of other app prototypes 10 Number of teams applying for in-depth support 78 Number of individual applicants to services 223 Teams that received in-depth support 21 Table 7: Number of people trained 223 Results Number of events 9 Highlights for mLab ECA Number of developers and entrepreneurs reached 1,578 Note: The percentage values represent linear comparisons across data for all four mLabs, with the mLab scoring highest for a given indicator receiving 100%. leverage the startups’ resources and mLab also draws on the services of strengths for each other. The mLab some of the incubated startups direct- selected a mix of service-based start- ly, for example, for design tasks, the ups (active in coding, design, training, mLab Studio app,67 or outreach to the digital media and gamification, and so tech community. on) and product-based startups (im- plementing a one-of-a-kind idea and Like other mLabs and mHubs, mLab business model). Startups contract Southern Africa has built a wide part- Appchemy or Innoprenez for app devel- nership network through ad hoc ini- opment when they are short-staffed, tiatives. Large tech companies (Nokia, draw on Jatamobile for design help, or Microsoft, Blackberry, Qualcomm, collaborate with Geekulcha if they need MTN), universities, the mLab’s incu- to reach out to South Africa’s developer batees (Geekulcha, Appchemy), and community. Even the more advanced many other organizations active in startups (such as Afroes or GoMetro), innovation and economic development which are largely active outside of (The Innovation Hub, AppCampus, Pretoria, consult with peers at the open.NASA, Ungana Afrika,68 SAIS,69 mLab, for instance, to access design STIFIMO,70 TIA,71 SAINe72) have part- 2 86 expertise or seek coding feedback. The nered with the mLab—often informally Overall Revenue generated reach by startups 100% 80% Investments Events 60% raised by startups 40% 20% 0% Applications Jobs created for support & supported Training Startups & startup teams created In depth App customer team support traction Figure 14: App innovation, (apps Comparative monetized, brought to Results for market, prototyped) mLab Southern Africa Note: The percentage values represent linear comparisons across data for all four mLabs, with the mLab scoring highest for a given indicator receiving 100%. and without contributing financially— help startups and early-stage innova- to run dozens of outreach campaigns, tors to engage in knowledge exchange networking events, hackathons, work- and opportunities for market extension. shops, trainings, and so forth. Box 7: Noted startups and app products supported by mLab Southern Africa Results mLab Southern Africa has recently Challenges shown good results and great potential The mLab has made remarkable prog- to serve developers, innovators, start- ress in recent months, but it still faces ups, and the South African mobile in- several challenges. The most fre- novation ecosystem as a whole. The re- quently mentioned one is location. The sults (see table 7) reflect that the mLab mLab is hosted inside The Innovation has supported only a few startups, but Hub Science Park. While proximity to it seems to have supported them well. offices of CSIR, The Innovation Hub, DST, and few tech businesses in the It should also be noted that these indi- park has been advantageous for some cators hide the substantial ecosystem startups (see section on Business impact that the mLab is having. The Model), the location is far from areas “vertical model” (see Sustainability and where student tech communities and Next Steps) of setting up a lean mobile ICT startup clusters have formed, so tech program for ICT incubators across that “organic,” regular, and serendipi- South Africa (and later the region) is an tous interaction with these stakeholder 2 87 ambitious but promising approach to groups is difficult. Public transport is patchy and highly inconvenient, idea and raw talent, even if this im- and interviewees reported that many plies a lower success rate and lower mLab members spend several hours probability that client entrepreneurs commuting. Entrepreneurs suggested can afford to pay for the support. that shuttle buses from central loca- tions and university campuses, onsite Third, like other mLabs, mLab Southern accommodation, or a satellite office Africa also struggled to get off the in Hatfield (Pretoria) could ameliorate ground and establish smooth opera- the situation. In its current location, tions within a short implementation the mLab cannot function as an open timeframe and given administrative space for cocreation and informal net- challenges and limited core funding. working, which is a prevalent model The original leader for the mLab project among technology innovation hubs. abruptly left to pursue another career opportunity. According to interviewees, The second challenge is the provision the first appointed mLab manager then of hands-on and in-depth mentoring overly emphasized revenue generation and coaching. The mLab tends to for the mLab and sought direct financial serve young entrepreneurs with little contributions from partners, which they experience, and often much support were not prepared to make, given the is required to turn skills and talent mLab’s initially lacking proof of concept into sustainable startups. Interviewed and low brand value. Importantly, the stakeholders identified the need for a lack of realistic targets meant that it was more stringent selection process to harder for the involved parties to make exclude those that are not committed the case that the current model was to their startup project. They also ad- not working; the manager was, at first, vocated that the mLab provide daily, able to point to completion of all goals in-depth coaching by well-matched specified in the original agreement. The mentors (including on business second and current manager was ap- modeling, go-to-market strategies, pointed in mid-2012. He swiftly ran an project management, and intellectu- outreach and sponsorship campaign, al property), regular mock pitching and implemented a host of ad hoc activ- to the immediate mLab community, ities, asking partners only for brand af- elaborate milestone setting and suc- filiation and in-kind contributions. Still, cess tracking, as well as follow-on the mLab continued to be overwhelmed support even beyond the incubation with setup activities (such as staff re- period.73 Several interviewees point- cruitment and compliance with admin- ed to Raizcorp’s thorough incubation istration and reporting requirements), support, although it was unclear if so that little time remained to consult the startups were aware of the modal- incubatees. Only in 2013 was the mLab ities and compensation that Raizcorp able to focus on a more thorough rollout requires in exchange for its services. of its incubation strategy. Yet, others held that the mLab should continue to start supporting entrepre- The final challenge concerns the role 88 neurs when they have just a business of infoDev and the World Bank for the mLab. Interviewees lamented a lack beyond with a mobile technology-spe- of clarity on the mission and goals of cific program. In mid-2013, it agreed infoDev and the World Bank. It was with Bandwidth Barn74 to establish proposed that infoDev use its under- an mLab site that would be colocated standing of the needs and challenges with the Cape Town incubator. For an of the mLab and its clients to broker interim period, the mLab leader will contacts and share lessons across the commute between the two sites and mLab and mHub network. One sug- manage both of them. This setup gives gestion was to send quarterly updates the Bandwidth Barn an additional and requests for feedback to all stake- resource for a sector that is in high holders, including to grassroots orga- demand while it bolsters the mLab’s nizations and donor representatives at presence in a highly frequented ICT local Finnish embassies. What is more, incubator and enables access to the infoDev was expected to build more Cape Town tech community. Next, the linkages to other World Bank units and mLab intends to spread its reach to activities in South Africa. other cities in South Africa through similar partnerships with the Durban Sustainability and Next Steps University of Technology and the gov- With its Pretoria site running at full ernment of the Eastern Cape Province. capacity, the mLab is in the process It is also in discussion with several of building out a “vertical model” to Raizcorp75 incubators. Most interview- complement existing entrepreneur- ees agreed that locations outside of the 89 ship support across South Africa and Pretoria-Johannesburg area and Cape 2013 2014 2015 2016 Expenses Office expenses and salaries 218,644 241,602 266,970 295,001 Travel 14,118 15,600 17,238 19,048 Sundries 4,235 4,680 5,171 5,714 Total expenses 236,997 261,882 289,379 319,764 Income Services and sponsorships Developer ecosystem outreachb 4,706 4,706 2,941 1,176 Training 1,176 2,353 2,353 2,353 App development work 4,706 5,882 9,412 11,765 Consulting 588 588 1,176 2,353 Royalties/equity — 1,176 23,529 — Grants infoDev grant 47,600 — — — Provincial/National government (partially pending approval)a,b 270,588 536,471 600,000 741,176 Table 8: Budget and Financial Total income 329,365 551,176 639,412 758,824 Projections for mLab Southern Financing gap/surplus 92,368 289,295 350,033 439,060 Africa 1,170,755 Total financing surplus until 2016 Note: All amounts in US$. Budget items are kept at a high level to avoid disclosure of sensitive information. Projections are based on rough estimates, as mLabs are currently in a transition phase at the end of grant funding from infoDev. a. Subject to approval from funders. b. Includes marginal revenue, that is, some expenses relating to the revenue item are not mentioned under expenses. Town hold tremendous untapped talent at the base of the pyramid (BOP). A and opportunity, and that mLab should more explicit BOP strategy could also pursue a “hub and spoke” approach be augmented by partners like Nokia with a mix of physical satellite offices, and Ungana Afrika, and it would fit events and competitions, as well as into DST’s mandate and national plan. virtual activities. Similarly, together with Dimension Data, mLab seeks to give developers To enhance the unique value that it privileged or subsidized access to can add, the mLab is in the process of cloud computing packages.77 Also the building access to unique resources for CSIR Meraka Institute’s intellectual startups. For example, it is discuss- property and expertise could be lever- ing with Balefyre how its proprietary aged more than in the past, in particu- integration platform enabling USSD lar, through its Mobi4D platform.78 As a service access through smartphones76 final value differentiator, the mLab in- can be made available to developers, tends to build on success stories such potentially opening opportunities to as GoMetro—for which mLab’s ties to pursue business models that target the public sector opened doors that are 90 users of feature and “dumb” phones usually closed for a small startup—and turn them into an explicit strategy: mLab could become a matchmaker for public agencies that seek complex, scalable app solutions and startups that implement them. mLab Southern Africa has also secured multiyear contributions from both pro- vincial and national government agen- cies, making it the most secure mLab in terms of sustainability and potential longevity, with a projected surplus for each year from 2013 to 2016. However, currently funded activities will large- ly be limited to Gauteng and some virtual support outside the province. Successful contracts with new pro- vincial governments will have similar limitations. Additional donor funding could help the mLab to achieve a more flexible and truly regional reach, for instance, by establishing satellite sites as part of the “vertical model,” or by running regional innovation events and competitions. The mLab is also seeking to further diversify its revenue streams. Over the coming years, management fees for facilitation of app development orders—especially from enterprise However, startups like Geekulcha and customers—are expected to grow Innoprenez prefer to “give back” to the substantially. Additional flows of in- mLab through in-kind contributions come will likely be generated from such as community outreach and commissioned training and occasional training. This suggests that the mLab consulting appointments. might have to offer differentiated suc- The mLab only began to experiment cess-sharing options based on the val- with equity schemes in 2013, which ue that it receives back from startups could lead to returns from liquidations during the incubation period. in the coming years. Several inter- viewed startups were open to giving up small equity shares (between 5 and 10 percent) in exchange for the service 91 value that they receive from the mLab. 92 KEY PARTNERS KEY ACTIVITIES VALUE PROPOSITIONS CUSTOMER CUSTOMER SEGMENTS RELATIONSHIPS • Enterprise Incubator Foundation • “Internship” program (interns work • Applied business & entrepreneurship • Student & graduate mobile skills,technical skills development • Niche, self selection of interesting developers on multiple app teams over few months) events and opportunities • Mentoring & coaching of internal • Team-building • Training partners (Microsoft & external teams Innovation Center, Armenian-Indian • Team leads • Confidence & enthusiasm • Personal, team leads at mLab Center of Excellence, Gyumri IT Center) • Subsidies for training programs, receive hands-on assistance intern referrals • Mobile app startup teams Student • Mentoring & coaching, & graduate mobile developers partnership opportunities • Personal, mLab coordinators • Ad hoc partnerships for grants, • Organization / contribution to give direct guidance events, & competitions (Microsoft innovation competitions, networking • Competitive exposure Innovation Center, Armenian Indian events (Regional App Contest, Center of Excellence, Nikita Mobile, m2Work hackathon, YAN hackathon) • Access to seed funding & grants mHub Georgia, mHub Moldova, regional coordinators in Ukraine and Georgia, AITT (Moldova), Gyumri • Student outreach The Business Model Canvas – mLab ECA Economic Development Foundation • Mentoring & coaching, (GEDF), Gyumri IT center (GITC), partnership opportunities Ministry of Economy fo RA, EDMC KEY RECOURCES CHANNELS (USAID), Cherie Blair Foundation for Women, Sourcio CJSC, MyNews NGO, SEUA, UITE) • Wide network & close contacts in tech, • University outreach incubation, economic development • Calls for participation in • Experience in implementing competitions & events through • Universities & technoloy centers tech incubation projects social media, mailings • Armenian government (direct pay- • Established technical training programs ment for renovation, equipment, • Personal interaction at mLab & furniture) • Speakers, coaches, sponsorships • Personal interaction upon request • Student body with basic skills • Modern facility, co-located with Microsoft Innovation Center & Yerevan State Univ. COST STRUCTURE REVENUE STREAMS • Main staff • App marketing • Provincial & national government • infoDev grant • Events • Office maintenance and other • App development & consulting • Developer ecosystem outreach • Trainings (Nokia) Note: To access the full canvas for mLab ECA, go to https://bmfiddle.com/f/#/H49t2. Each partner relationships, channels, and value proposition elements have the according color when they mainly or group of partners has a specific color; activities and resources that mainly rely on a particular relate to the customer segment in question. Activities and resources as well as cost and revenue partner group have the same color. Also each customer segment has a distinct color, and streams that only relate to the infoDev grant have their own color. Building teams, products, and businesses, for the long-term support of a fledgling startup ecosystem mLab ECA Total infoDev grant funding $585,000 Funding for technical assistance $426,000 Base city, country Yerevan, Armenia infoDev grant start date May 2, 2011 First activities launched July 11, 2011 Official launch September 12, 2011 Launch in expanded facilities June 13, 2013 Host organization Enterprise Incubator Foundation Key partners Ministry of Economy of Armenia, Microsoft nnovation Center Armenia, Armenian-Indian Center for Excellence in ICT, Gyumri IT Center, Nikita Mobile, regional partners (mHubs in Moldova and Georgia, partners in Ukraine, and Georgia), AITT (Moldova), Gyumri Economic Development Foundation, EDMC (USAID) Targets as per grant agreement • Organization of several developer events, competitions, and challenges, including a regional challenge in collaboration with mHubs of the ECA region and other partners covering 11 countries • 200 students trained in mobile app development, business & entrepreneurship training • 15 incubatees with products brought to market • At least 10 startups hosted within the mLab Website www.mlabeca.com Table 9: Social media and other web resources @mLabECA, https://www.facebook.com/mLabECA, http://www.youtube.com/channel/ Basic Data UChY1MstyR7Cz7h4SJwsNuFg for mLab for mLab ECA. Note: Funding amounts are based on actual expenses as of September 2013. Amounts are rounded to thousands. Business Model join an internship program, during mLab ECA,79 in its first year and a half of which they are encouraged to try out operations, has focused on talent devel- mobile entrepreneurship while honing opment and team building. mLab ECA their skills and interests and forming has taken a long-term approach: the lasting connections with peers. The program does not envisage simply cre- mLab provides applied product devel- ating as many startups as possible, as opment skills for all interns and direct quickly as possible, but rather aims to mentorship for team leaders. It also produce talented serial entrepreneurs accepts existing, fledgling startups for Figure 15: committed to building an Armenian acceleration. The Armenian developers (opposite page) startup ecosystem from the ground up. and entrepreneurs interviewed were Business This is reflected in the mLab’s value probably the most motivated and en- Model Canvas for mLab ESA proposition. Young graduates with no thusiastic of all, reflecting mLab ECA’s 93 previous entrepreneurship experience efforts to spur confidence and a sense of commitment in their clients. Through of Excellence in ICT84 (AICT), and the innovation competitions, the mLab Gyumri Information Technology Center.85 exposes them to competitive pressure Innovation competitions included the and pushes them to sharpen their app m2Work and YAN hackathons. Most products. The mLab also helps teams recently, the mLab ran a Regional to secure grants and small seed invest- Mobile App Contest, partnering with ments, for instance, from EDMC80 or innovation hubs, incubators, and tech CRDF Global.81 A secondary customer community organizations in eight segment is made up of existing startups countries to multiply outreach. Awards that approach mLab to receive advice went to one winner from each country and connections. and one overall winner. The mLab’s long-term vision is also in Results line with the mission of the Enterprise mLab ECA has made headway in cata- Incubator Foundation (EIF), the organi- lyzing the Armenian tech startup eco- zation that set up the mLab and that is system, but it is not yet a powerhouse still its main partner. EIF opens doors of startup creation and economic for the mLab and its clients to its vast impact. So far, it can point to the cre- network of actors in technology incu- ation of five startups and ten advanced bation and economic development. startup teams. The relatively limited The mLab’s goals—in particular with results are due to the mLab’s long- regard to skills development—are also term approach and relatively short im- complementary to the Armenian gov- plementation period. Unsurprisingly, ernment’s efforts to position the coun- this also led to modest numbers for try as a stronghold for technology in- job creation, revenue generation, and novation in the region. The government raised investments by startups and sees the mLab as a capacity-building startup teams (see table 10). hub for the country and the region, and has supported it by financing the On the upside, apps produced by devel- renovation and equipment of a new opers in the mLab’s internship program facility.82 The facility is colocated with have been wildly successful in regional the Microsoft Innovation Center and and global app stores, with a total num- the Armenian National Engineering ber of downloads of about 1.7 million. Laboratory on the campus of the State Together with more impressive results Engineering University of Armenia. for output and immediate outcome In turn, interviewed government offi- indicators (for instance, 242 trained cials assured that government would individuals and 131 app prototypes refrain from actively engaging in the developed), mLab ECA seems to tap mLab’s operations. into and nurture tremendous potential, in particular with regard to Armenian Training programs were funded by mobile app developers’ technical so- mLab and conducted in collabora- phistication and design skills. Also the tion with the Microsoft Innovation local government’s now solid buy-in 94 Center,83 the Armenian-Indian Center can be seen as a positive result; the mLab appears to have delivered at least a proof-of-concept that can be the basis for further investments. Slowly but surely, the mLab’s entrepre- neurs are improving their standing and are beginning to turn their ideas into “serious” companies. For instance, one of the entrepreneurs came fifth out of 40 in a Dragon’s Den at infoDev’s Global Forum, 86 an international pitching competition. Other entrepre- neurs have achieved recognition and exposure through Microsoft’s Imagine Cup,87 UITE’s Digitec Business Forum88 (one of Armenia’s main conferences for the digital economy), or Armenia’s startup cup.89 Indicator Result as of June 2013 Revenues generated by startups $55,100 External investments raised by startups $128,700 Direct jobs created by startups 37 Startups created 5 Solid startup teams that have not registered a business 10 Consumer customer traction: Number of app downloads/users/subscribers 1,700,000 Organizational customer traction: Number of enterprise and public agency customers 209 Apps brought to market 117 Number of other app prototypes 131 Number of teams applying for in-depth support 216 Number of individual applicants to services 242 Teams that received in-depth support 57 Number of people trained 242 Number of events 27 Table 10: Number of developers and entrepreneurs reached 1,000 Results Highlights for Number of developers and entrepreneurs reached 400 mLab ECA Note: As reported by mLab manager. Including estimates, proxies, and underreporting due to missing data 95 Overall Revenue generated reach by startups 100% 80% Investments Events 60% raised by startups 40% 20% 0% Applications Jobs created for support & supported Training Startups & startup teams created In depth App customer team support traction Figure 16: App innovation, (apps Comparative monetized, brought to Results for market, prototyped) mLab ECA Note: The percentage values represent linear comparisons across data for all four mLabs, with the mLab scoring highest for a given indicator receiving 100%. Noted app products supported by mLab ECA Box 8: Noted app products supported by mLab ECA Note: Websites at http://www.globalforum2013.co.za/downloads/presentation/Day%202/Dragons’%20Den%20-%20 Day%202/MicroForester%20Presentation.ppt (ppt presentation); http://www.edmc.am/clean-h2o-project-wins-the- armenian-round-of-imagine-cup-2012/; http://www.mlabeca.com/eng/mobile_applications/windows-phone/the-gar- dener-2/; http://www.mlabeca.com/eng/mobile_applications/ios/paint-in/. A more comprehensive list of supported startups and teams can be found in the appendix. Detailed case studies of these and other startups will be featured in an upcoming infoDev publication, commissioned to the private sector development evaluation and research agency 96 CAD (http://cad.globalcad.org/en/que-es-cad/). Challenges been tough before the launch of a new The major challenge that mLab faced facility where the mLab is now able to was an insufficient seed budget to hold events and workshops. build up sufficient infrastructure and For mLab ECA, more than for other scale. The grant size of $585,000 was mLabs, it would have added value to seen as insufficient to establish the collaborate with other mLabs, giving mLab as an entity that could tackle Armenian entrepreneurs exposure to the more fundamental gaps in the larger markets and helping them to ecosystem and fulfill the broad and internationalize their businesses. Yet, long-term vision that it felt was the mLabs had little time beyond their most purposeful. Interviewees point- immediate local deliverables, so that ed out that mLab ECA could have some of mLab ECA’s efforts to reach out benefitted from more hands-on and did not result in concrete joint projects. swift help from infoDev, especially in the setup phase. Although govern- The mLab also had to deal with other ment is supportive of the mLab, it was challenging external factors. In partic- hard to raise substantial coinvest- ular, several interviewees pointed out ments, since the proof-of-concept that entrepreneurs face taxation and stage could not be reached quickly other legal barriers when they want enough, and government funding to formally register a startup. Much above $500,000 is usually subject to of the talent base relocates abroad; approvals at the ministerial level or many interviewees lamented “brain the prime minister’s office, where drain” as an important problem. officials have become more aware of Interestingly, the relatively advanced the mLab only recently. state of the traditional software and IT industry can also be a problem for Similarly, the mLab realized that the business creation, as talented devel- implementation timeline of less than opers often find regular employment two years of operation was too short that is more lucrative than working at to create a portfolio of startups that a startup. As a result, the mLab has would include high-impact app start- found that it has to target young inno- ups and those with more complicated vators, who have not yet established business models. Despite consider- themselves, and that it has to address able growth over the past few years, more fundamental issues, such as the Armenia still cannot boast a vibrant lack of entrepreneurial culture and tech community. Pockets of activity skills, before founding a startup will at universities and in the blogosphere become a path that many clients are are disintegrated, and few events comfortable to pursue. are taking place that would give tech entrepreneurs a chance to network Sustainability and Next Steps and learn. Moreover, the creation of mLab ECA’s management has found strong, specialized subcommunities that the Armenian and regional mar- that could bring to bear more sophis- kets hold limited potential for incuba- 97 ticated apps and business models had tors focused on mobile app startups to 2013 2014 2015 2016 Expenses Main staff 49,263 110,634 112,338 114,211 Office maintenance and other 16,000 32,150 32,150 32,150 App marketing 1,500 12,000 12,000 12,000 Total expenses 66,763 154,784 156,488 158,361 Income Apps and products 7,100 15,000 23,000 37,375 Project grants 9,000 20,000 22,000 26,400 Educational component 9,073 36,000 39,600 47,520 Events 10,000 10,000 11,000 13,200 Startups (membership fees, success sharing) 4,390 10,000 11,000 13,200 Table 11: Total income 39,563 91,000 106,600 137,695 Budget and Financing gap (27,200) (63,784) (49,888) (20,666) Financial Projections for Total financing gap until 2016 (161,538) mLab ECA Note: All amounts in US$. Budget items are kept at a high level to avoid disclosure of sensitive information. Projections are based on rough estimates, as mLabs are currently in a transition phase at the end of grant funding from infoDev. a. Excludes basic facilities and physical infrastructure, which was provided in-kind. be profitable. In particular, the “ early At this point, the mLab is leveraging stage” teams that the mLab nurtures funds and resources from its parent are neither in the position to pay for ser- organization, EIF, and it benefits from vices, nor to quickly increase their val- the government’s support for the new uation to a level where the mLab could facility. The mLab keeps the revenue make substantial returns on (planned) that is generated through apps created equity investments. The management during internship projects, but this rev- estimates that, under current condi- enue stream is, to date, far from cov- tions, pathways towards profitability ering the mLab’s costs. Overall mLab would require a drastic and unwanted revenues other than grants cover about mission change away from a broad half of its expenses. ecosystem-oriented approach towards, for instance, evolving the mLab into a In the future, the mLab intends to seek product development agency for “killer additional resources and funds from lo- apps,” which is highly selective and only cal government and donors. In addition, focuses on a few highly promising or the mLab wants to function as a kind of already proven apps. clearing house for app development, claiming fees in exchange for broker- Consequently, while mLab ECA was age of grants and seed investments able to take the first steps towards di- for entrepreneurs: often, investors and versifying revenue streams, it still de- grant programs call for mobile tech in- 98 pends on donor funding and subsidies. novations but mobile developer teams by themselves do not have the institu- mLab ECA is keen to further extend tional capacity to be investable or they its reach beyond Yerevan, at the rural, are not eligible to apply for support. regional, and global level. Locally, the In such cases, the mLab could man- mLab has established a representation age the financial relationship, helping in Gyumri and is planning to partner teams administratively while keeping with a technology center in Kapan in a success fee. The mLab also wants southeast Armenia. Regionally, the to monetize its training programs, mLab is pushing for a more formalized potentially by introducing a layered collaboration of innovation hubs and pricing scheme that distinguishes be- incubators, for example, in the form tween corporate training and training of an association or a “hub and spoke” for individual students. Moreover, the approach led by EIF and the mLab. mLab looks to leverage its access to Globally, the mLab mainly intends the developer community, its brand, to build out partnerships with other and organizational capacity to attract mLabs and benefit from Armenia’s sponsorship for events as another prof- Representation Office in Silicon Valley. it potential. Finally, the mLab has taken The mLab could also benefit from EIF’s equity positions in three startups and efforts to set up an IBM Research Center intends to grow this approach, opening in the same building and the launch of a up the potential of income from equity venture capital fund for Armenia.90 liquidations in the more distant future. 99 100 KEY PARTNERS KEY ACTIVITIES VALUE PROPOSITIONS CUSTOMER CUSTOMER SEGMENTS RELATIONSHIPS • Events and competitions • Events, competitions, workshops, • Basic skills development • Student & graduate mobile (Nokia, Samsung, UNICEF, networking meet ups and developers • Niche, self selection of StartMeUp, AiTi (MoMo), Qualcomm, demo days • Networking platform: interesting events and opportunities FPT Software, App Campus) awareness & access to partners & job opportunities • Mobile app startup teams • Tech ecosystem building • Personal interaction of mLab • Ecosystem outreach coordinators, coaches, & mentors (Tech in Asia, Action.vn, Barcamp • Mentoring & coaching, partnership opportunities • Event and workshop customers Saigon) • Technical outsourcing (Samsung, Nokia, UNICEF) for incubatees • Pitching opportunities • Personal interaction with reps for developer outreach • Outsourcing • Mentoring & coaching • Outsourcing clients (Free Range Technologies, Technologic Arts) of incubatees • Access to developer community • Mentoring KEY RECOURCES CHANNELS (DFDL, HSC, IDGVV, CyberAgent Ventures, FPT University) • Technical expertise • Mailing lists • Motivated group of developers, • “Multipliers” at universities, campaigns The Business Model Canvas – mLab East Asia • Saigon High Technology app entrepreneurs • Face-to-face outreach workshops Park, Saigon High Tech Business at universities Incubator • Outreach coordinator • Personal interaction after events • Local consultant • Direct outreach around events & opportunities COST STRUCTURE REVENUE STREAMS • Projects (trainings, workshops, etc.) • Outsourcing project management fees • Staff • Service fees for events and competitions • Operating Expenses • infoDev grant • Equipment & software Note: To access the full canvas for mLab East Asia, go to https://bmfiddle.com/f/#/Mhz29. Each and relationships, channels, and value proposition elements have the according color when they partner or group of partners has a specific color; activities and resources that mainly rely on a mainly relate to the customer segment in question. Activities and resources as well as cost and particular partner group have the same color. Also each customer segment has a distinct color, revenue streams that only relate to the infoDev grant have their own color. Events, competitions, and lightweight acceleration for greater reach mLab East Asia Total infoDev grant funding $230,000 Funding for technical assistance $480,000 Base city, country Ho Chi Minh City, Vietnam infoDev grant start date December 19, 2011 First activities launched September 14, 2012 Official launch September 17, 2012 Consortium members Saigon High Technology Park (SHTP) and Saigon High Tech Business Incubator (SHBI) (lead), Vietnam National University, FPT University, Elcom Key partners Nokia, Blackberry, Samsung, UNICEF, AITI, StartMeUp, Vietnam Youth Entrepreneurs, Barcamp Saigon, Saigon Hub Targets as per grant agreement • Development of short training courses with at least 50 students trained in mobile app development • Development of vocational training program with at least 20 students enrolled to the program • Ten incubatees selected and placed within the mLab to benefit from the incubation program services • Generation of between 8-10 mobile applications by 2013 Website http://www.mlab.vn Social media and other web resources @mLabEastAsia, https://www.facebook.com/mlab.eastasia, Table 12: , http://www.youtube.com/user/mLabEastAsia, Basic data for mLab http://mic.mlab.vn/en/ East Asia Note: Funding amounts are based on actual expenses as of September 2013. Amounts are rounded to thousands. Business Model ideation and prototyping challenges After its launch in September 2012, to competitions focused on startup mLab East Asia focused on events and creation, which were embedded in a competitions to reach app developers multistage “lightweight” acceleration and early-stage entrepreneurs. The scheme for ten startup teams. core of its value proposition was to provide developer teams with skills Originally, the mLab focused mainly on Figure 17: development as well as opportunities students. It reached out to about ten (opposite page) universities in the wider Ho Chi Minh Business to showcase app projects and develop Model Canvas new ideas for vibrant mobile content City catchment area. It employed tra- for mLab East Asia markets in Vietnam and East Asia. In ditional and multimedia outreach cam- 101 2013, the mLab started to shift from paigns combining an online presence, social media, tech blogs, conventional The most promising teams and en- media advertisements, and printed trepreneurs from these events were brochures to advertise events and screened and invited to join the mLab innovation contests to technology incubation program. In June 2013, nine students and engineers. The m2Work teams signed incubator agreements hackathon and the Vietnamese leg and agreed to offer mLab 10% equity of the regional Blackberry Jam Hack in their startups in exchange for men- were highlights of 2012. The mLab also torship, training, networking oppor- ran an Android workshop as well as tunities, business support, and seed Windows and Asha phone workshops, capital ranging from $2,000 to $5,000. together with Nokia. The leader of each team joined a week- long bootcamp with pitch training, or- The mLab’s activities slowed down ganized by the local partner Viet Youth in late 2012 when the original mLab Entrepreneurs in collaboration with manager resigned. In response, the Stanford University and Guy Kawasaki. Saigon High Technology Park (SHTP), as the consortium leader, assigned the Results director of its existing incubator, the Saigon High Tech Business Incubator mLab East Asia has certainly not (SHBI), as interim mLab manager. achieved everything that would have Later, infoDev hired a local consultant been possible in the quickly growing as a support resource. mobile innovation ecosystems of Ho Chi Minh City and Vietnam. Yet, mLab has From March 2013 onwards, the mLab recently made progress. Most signifi- launched several projects in quick cantly, it was able to identify a portfolio succession. At the core was the mLab’s of talented startup teams—two of the Mobile Innovation Challenge,91 which mLab’s teams went on to become the awarded a total of $14,400 to five winner and runner up in DEMO ASEAN winners chosen from a pool of fifteen 2013. The mLab also added new reve- finalist teams. 106 teams had applied. nue streams as it established its brand. The challenge followed a format Recent developments have also shown comparable to a smaller version of that a variety of partner organizations mLab East Africa’s Pivot East, offering are willing to support the mLab. It is workshops for Finalists to prepare widely seen as a promising and desired them for a pitching contest. The mLab program to alleviate the current lack of launched the Smart TV App Challenge structured, continuous startup support. in collaboration with Samsung, which Yet, in June 2013—ten months after culminated in an award ceremony at its launch—the mLab was still very the end of October 2013. It went on to much in the early rollout phase. A lot of host a hackathon for UNICEF coupled momentum was created, but only the with a four-month program to prepare coming months and years will show if finalists for a pitching contest with the mLab and its consortium are able to $15,000 in prize money. Hundreds of formalize partnerships or secure addi- trainees joined additional workshops tional donor funding in order to sustain 102 on Samsung and Windows technology. and institutionalize operations. Overall Revenue generated reach by startups 100% 80% Invetsments Events 60% raised by startups 40% 20% 0% Applications Jobs created for support and supported Training Startups & startup teams created In depth App customer team support traction App innovation, (apps monetized, brought to Figure 18: market, prototyped) Comparative Results for Note: The percentage values represent linear comparisons across data for all four mLabs, with the mLab scoring mLab East Asia highest for a given indicator receiving 100%. Indicator Result as of June 2013 Revenues generated by startups $12,600 External investments raised by startups $64,000 Direct jobs created by startups 54 Startups created 2 Solid startup teams that have not registered a business 11 Apps brought to market 17 Number of app prototypes 183 Number of teams applying for in-depth support through mLab 245 Number of individual applicants to services 697 Teams that received in-depth support 19 Table 13: Number of people trained 852 Results Number of physical, face-to-face events organized 20 Highlights for mLab Number of developers and entrepreneurs reached 10,000 East Asia Note: As reported by mLab manager. Including estimates, proxies, and underreporting due to missing data 103 Owing to the early focus on ideation and mLab already had a late start—the orig- skills development for students, gov- inal schedule had foreseen activities to ernance issues, and several periods of start at the beginning of 2012 and not inactivity, only the last quarter of oper- in September. Soon after the launch, ations in mid-2013 hinted at the mLab’s differences in perspective between the strong potential. The modest results original mLab management and the ought to be seen in the context of a short, consortium led to—finally unassail- challenge-ridden implementation time- able—concerns over the completion of line (table 13). A handful of startups deliverables for infoDev and consortium were registered and the first external representatives. From end-2012 to investments were raised. Notably, the mid-2013, SHTP and SHBI took on the demand and potential for the mLab’s interim management, but were not able services is vast, as is reflected in the to commit the time and resources nec- number of informal partnerships, the essary to reinvigorate the project. number of applications for support, and In early 2012, the recruitment of a local the mLab’s overall reach. infoDev consultant helped to jumpstart Challenges an array of activities, but following the end of the infoDev grant in June 2013, mLab East Asia faced several imple- the mLab entered another difficult mentation challenges. First, the mLab’s transition period. The interim mLab location in the SHTP science park in Ho manager left SHBI, leaving a vacuum Chi Minh City’s District 9 was not a via- of institutional knowledge of how to ble gathering place for developers and effectively engage in existing and new entrepreneurs. The space is a thirty to activities. SHBI has not yet been able to sixty minutes’ drive from the city center put a consistent management structure depending on traffic (outside of rush in place, and a concept to secure cocre- hours); public transport is available but ation space downtown was not followed not convenient for most clients. Clients through. infoDev is actively working expressed that they have no incentive with SHTP and SHBI, as well as repre- to use mLab’s physical space or testing sentatives from the World Bank country lab on an ongoing basis, and preferred office to resolve these issues. a coworking space in a more central, downtown area. Instead, they were sat- Sustainability and Next Steps isfied to continue working from home, Most recently, the mLab has shown school, or the numerous cafés that offer potential to generate revenue and sub- free wireless broadband. As a result, stantive in-kind support from the pri- the mLab space in the high-tech park vate sector, securing outside support has consistently been underutilized and that ensured continuous activity until proven difficult to populate. the end of 2013. Contributions came in the form of coordination of outreach Second, operations were slowed down at campaigns and competitions, training several points in time. Owing to bureau- and mentorship of startups in the cratic hurdles in the interplay between incubation program, as well as orga- 104 the World Bank and the consortium, the nization of investor networking events Noted startups and app products supported by mLab East Asia Box 9: Noted startups and app products supported by mLab East Asia Note: Websites at http://mic.mlab.vn/en/uncategorized/english-ismartbike/; http://www.lifebox.vn/; http://watermel- on-studio.com/. A more comprehensive list of supported startups and teams can be found in the appendix. Detailed case studies of these and other startups will be featured in an upcoming infoDev publication, commissioned to the private sector development evaluation and research agency CAD (http://cad.globalcad.org/en/que-es-cad/). 105 and informal meet-ups. A number of workshops and networking events were planned, culmi- nating in a demo day networking event where incubatees could meet investors form the region. The mLab training (Nokia and UNICEF Mobile also launched the mDevNetwork to Hackathon). It expects to generate respond to the demand for technical further revenue from commission outsourcing and to provide incubatees fees for technical outsourcing proj- with an opportunity to earn their own ects, annual partnerships with other seed capital. private-sector organizations, con- sulting projects, and funding from From April to October 2013, mLab development agency donors and local East Asia was able to secure consult- government agencies. ing income from the private sector (Samsung Smart App TV campaign) and sponsorship for incubation and 2013 2014 2015 2016 Expenses Consultancy services (staff) 18,254 27,381 41,072 61,607 Projects 83,063 108,063 133,063 158,063 Equipment and software 62,642 17,642 27,642 37,642 Operating expenses 66,261 89,452 107,343 123,444 Total expenses 230,220 242,538 309,119 380,756 Income Sponsorship 8,553 75,000 90,000 100,000 Consulting 15,353 46,059 76,765 107,471 Training 4,971 6,711 9,060 6,711 Outsourcing 10,000 25,000 50,000 75,000 infoDev grant 153,000 Table 14: Budget and Total income 191,877 152,770 225,825 289,182 Financial Projections Financing Gap (38,343) (89,768) (83,294) (91,574) for mLab Total financing gap until 2016 (302,979) East Asia Note: All amounts in US$. Budget items are kept at a high level to avoid disclosure of sensitive information. Projections are based on rough estimates, as mLabs are currently in a transition phase at the end of grant funding from infoDev. 106 mLab South Asia proposal and frequently interacted with infoDev. A failed project, but many lessons This abrupt shift called the substance learned of all original agreements, including the winning proposal, into question, A planned fifth mLab in the South and it required substantial time and Asia region was never implemented. effort to establish a new relationship The following section will not analyze and common implementation agenda. this mLab’s proposed business mod- The new leadership then pointed to el, but briefly review the background previous procedural errors. Clarifying and then go on to derive lessons for how to resolve the errors that had been the setup of future mLabs and other made took a long time, and in fact some innovation hubs. of them are not resolved at the time of this writing. In sum, this meant that Background deadlines were not met and substantial At the end of 2010, a panel of infoDev transaction cost were incurred—both and World Bank staff (in coordination on infoDev’s and PSEB’s part. with Nokia representatives) selected the proposal submitted to an open In June 2013, the mLab grant expired. call by a consortium in Pakistan led by From infoDev’s view, any future activ- the Pakistan Software Export Board ity should avoid the mistakes made (PSEB). The bid was selected specif- and follow the lessons outlined in this ically because of the complementary report, in particular, a more hands-on strengths of the institutions participat- market scoping assessment and more ing in the consortium. PSEB already detailed design of the project by infoDev. had good relations to private-sector Lessons actors such as Nokia. When an mLab is implemented by a What followed was a history of gover- government-affiliated agency in collab- nance and leadership issues and en- oration with infoDev, a longer timeline suing delays. Changes at PSEB twice is needed. came at most inopportune times: shortly after the first tranche of the The timeline for mLab South Asia of grant had been disbursed in February about two years for the entire grant ad- 2012 and shortly after a revised im- ministration and implementation would plementation plan had been decided have been hard to meet under most on in August 2012. Naturally, differ- circumstances given the organizations ent leaderships had different priori- involved. After all, grantee deliverables ties as to the best way forward for the were to be completed by June 30, 2013, mLab. The most important challenge less than 17 months after the first was the exit of the PSEB’s executive tranche had been disbursed to PSEB. If director originally responsible for the grant preparations had been done the mLab, who had led the winning according to the official procedures, 107 the failure of the project might have What was missing in the case of mLab been possible to avert, but even then, South Asia was continuity of vision and implementation would have likely been commitments. Moreover, it was difficult slower than for other mLabs. The tight for infoDev to understand to what extent schedule would probably have sufficed the government entities behind PSEB only in the scenario of an entirely wanted to have a more direct influence smooth rollout. on the mLab’s decision making. At the least, this meant another worry for in- It would probably be an overgeneral- foDev, as any government influence on ization to conclude that mLabs cannot an mLab’s operations would usually be be implemented by government-af- a concern, given the necessary market filiated agencies.92 Relying on small and private-sector orientation. grassroots organizations can bring its own set of challenges (such as Importantly, leadership changes also inadequate administrative capacity), implied smaller and bigger vision and and government backing can help to strategy changes. The original direc- ensure financial sustainability and tor of mLab South Asia had been the high-level partnership brokering, as driving force behind the consortium’s long as it does not get directly involved successful application and, to infoDev, in the mLab’s operations (see the mLab he seemed the right individual, display- Southern Africa case for example). But ing pragmatism, an understanding of a project design and implementation entrepreneurship ecosystem dynamics, timeline of three years does seem un- and private-sector experience. But realistic for such a proposition. when he became unavailable, each new director needed to start from scratch to An mLab’s consortium lead organiza- learn how an mLab could work within tion should be internally stable, nimble, the strategic priorities set for PSEB as and driven to support the innovation an organization. ecosystem by its mission and funda- mental goals. infoDev did assess the risk of po- tentially slow processes and deci- As with all consortia, the member or- sion making within the consortium. ganizations’ goals and incentives need However, given PSEB’s and other to be aligned with the mLab’s. But in consortium members’ professional particular, mLabs need to be immersed demeanor and convincing assuranc- in the tech startup culture, as they rely es, this risk was not seen as a fault on openness to and engagement with line. This implies that infoDev should innovation ecosystems and grassroots not only take a deeper look into the tech and startup communities. They composition of a consortium but also also need to be swift and up-to-date the internal dynamics of a consor- to gain the respect and trust of these tium leader—including any potential communities. Accordingly, lead organi- decision-making constraints and zations in mLab consortia need to have interlaced hierarchies. From the re- an inherent interest and experience in viewed project documentation, it is 108 tech entrepreneurship support clear that infoDev did ensure that the best bid received the grant. However, itself more to change management infoDev could have conducted a deep- when PSEB’s original mLab director er scoping assessment and possibly resigned abruptly. In hindsight it be- the preparation of a high-level con- came clear that infoDev should have cept for mLab South Asia. With better helped the incoming consortium rep- knowledge about the market require- resentatives to understand the basic ments and consortium organizations, concept behind mLabs, and to iden- infoDev might have made a better risk tify a viable model for PSEB and the assessment. other consortium organizations. Yet, infoDev’s team also expressed that it infoDev needs to have resources avail- was overstretched with other ongoing able to conduct responsive and adap- responsibilities and that sudden peaks tive change and conflict management in the need for intervention were hard when a grantee faces challenges. to accommodate. For future mLab In the case of mLab South Asia, infoDev grants, infoDev will consider setting had difficulties in anticipating and aside staffing and travel resources that resolving unexpected implementation can be used when there are signs that bottlenecks and conflicts. In particular, a grantee faces structural and admin- 109 infoDev found that it could have applied istrative challenges. 110 KEY PARTNERS KEY ACTIVITIES VALUE PROPOSITIONS CUSTOMER CUSTOMER SEGMENTS RELATIONSHIPS • YIPL • Regular meetings & gatherings • Guidance for teams in pre- • Mobile developers & tech startup phase • Personal / niche / community, entrepreneurs • Technical trainings mobile technologists see Mobile • Multi-stakeholder networking Nepal as go-to place, trusted, • Ad hoc partnerships for events & • Online community at platform: access to partners neutral, enthusiastic competitions (invididual coaches / • Community champions mobilenepal.net, app showcase & job opportunities mentors / judgest, F1Soft, Nepak Open Source Klub, CSIT Association • Newsletter for mailing list subscribers • Personal, YIPL provides of Nepal) • Building leader profile in guidance, resources, & community, career opportunities opportunities directly • Large ecosystem conference • Nepalese Young Entrepreneurs Forum (NYEF) • Pivot Nepal • Universities (focus on Tribhuvan University, Institute of Engineering) • Student outreach CHANNELS • infoDev • m2Work hackathon • Mobile Nepal Website • Google Groups The Business Model Canvas – Mobile Nepal (mHub Nepal) KEY RECOURCES • Mailing list • Rolodex • Computer / IT & business student body • Event organization capacity • Regular personal interaction • Meeting room with Wifi connectivity, • Community facilitator generator • Toolkits, advice • Multi-stakeholder participation • High-level business contacts COST STRUCTURE REVENUE STREAMS • Pivot Nepal • Face 2 Face Meetings • infoDev: m2Work hackathon contract • m2Work hackathon • Web Portal & Internet Connectivity • infoDev: Pivot Nepal grant • Human Resources • Equipments • infoDev: mHub grant • Trainings Note: To access the full canvas for Mobile Nepal, go to https://bmfiddle.com/f/#/nYNv2. Each and relationships, channels, and value proposition elements have the according color when they mainly partner or group of partners has a specific color; activities and resources that mainly rely on a relate to the customer segment in question. Activities and resources as well as cost and revenue particular partner group have the same color. Also each customer segment has a distinct color, streams that only relate to the infoDev grant have their own color. Nepal’s aspiring mobile tech community, built from the ground up Mobile Nepal (mHub Nepal) Total infoDev grant funding $61,000 infoDev contract-based funding $30,000 Funding for technical assistance $6,000 Base city, country Kathmandu, Nepal infoDev grant start date July 1, 2011 First activities launched July 1, 2011 Official launch December 12, 2011 Host organization Young Innovations Pvt. Ltd. Key partners Nepalese Young Entrepreneurs Forum, Nepal Open Source Klub, F1Soft, Tribhuvan University Institute of Engineering, CSIT Association of Nepal, International Center for Integrated Mountain Development Targets as per grant agreement • Hosting social events, typically on a monthly basis, to promote collaboration and innovation in the mobile sphere • Organizing Pivot Nepal competition, including for fifteen selected applicants: mentoring; refinement of idea and development of prototypes; pitching mentorship Website http://mobilenepal.net/ Social media and other web resources @MobileNepal, https://www.facebook.com/pages/ Mobile-Social-Networking-Nepal/169545916438167 http://www.youtube.com/user/mobilenepal, Table 15: http://pivot.mobilenepal.net/, Basic Data for Mobile Nepal http://www.younginnovations.com.np/ (mHub Nepal) Note: Funding amounts are based on actual expenses as of September 2013. Amounts are rounded to thousands. Business Model opportunities, training, and—maybe Mobile Social Networking Nepal—or most importantly—motivation and simply Mobile Nepal—prides itself confidence. Young Innovations,93 the in having established the first tech organization behind Mobile Nepal, community on mobile apps in Nepal. has become a trusted and known en- In a little over two years, it has be- tity in the innovation ecosystem. The come the go-to place for “everything organizers of Mobile Nepal targeted Figure 19: mobile” in the country. a handful of community champions, a (opposite page) step critical to the community’s fast Business For app developers and early-stage growth and stability. The champions Model Canvas entrepreneurs, Mobile Nepal rep- quickly started to independently for Mobile Nepal (mHub Nepal) resents the platform where they drive activities such as workshops on convene to receive networking specific technical topics, keeping 111 participation alive in the virtual space doors to successful entrepreneurs and on mobilenepal.net. In addition, the high-level business circles. Pivot Nepal project leaders reached out to several also gave occasion for the organizers universities in the Kathmandu area to to build informal connections to a make students aware of opportunities wider group of stakeholders, including in mobile app entrepreneurship. About mentors, coaches, and judges. 25 to 50 community members now form the core, but the total reach has risen These activity highlights have not only quickly to almost 1,000 developers and solidified Young Innovations’ brand and entrepreneurs. In mid-2013, the Mobile credibility, but also its capacity to work Monday Kathmandu chapter was born as with donor organizations. This capacity has been instrumental in building ties be- the latest achievement of the community. yond the tech community to stakeholders Three highlights stood out. Following from government, banks, business asso- the start of community building shortly ciations, and international development after the grant disbursement in mid- professionals. Young Innovation was able 2011, Mobile Nepal hosted the Mobile to participate in Open Data Nepal99 and Ecosystem Forum,94 where Mobile Development Check,100 and it also ran the Nepal brought together a wide array Violence Against Women hackathon101 to- of high-level representatives from all gether with a branch of the World Bank’s relevant stakeholder groups in Nepal’s social development unit and the local IFC fledgling mobile innovation ecosystem. office. These projects open opportunities In 2012, Mobile Nepal not only hosted for app developers to gain experience its own leg of the m2Work hackathon,95 and income, which are critical in view of but also facilitated the event globally96 in still limited market opportunities. collaboration with the four operational mLabs. More recently, in April and May Results 2013, the Pivot Nepal97 competition was In sum, Mobile Nepal has sowed the the group’s attempt at a “mini-accel- seeds of a fledgling mobile innovation eration” program for startup teams, ecosystem by growing a core communi- based on the methodology of Pivot ty. At the same time, it has to be noted East, mLab East Africa’s mobile app that the low readiness of the mobile startup competition. Young Innovations innovation ecosystem did not allow for partnered with the Nepalese Young substantial startup creation and sus- 112 Entrepreneurs Forum, which opened tained entrepreneurial activity. Indicator Result as of June 2013 Direct jobs created by startups 13 Startups created 1 Solid startup teams that have not registered a business 4 Consumer customer traction: Number of app downloads/users/subscribers 600 Organizational customer traction: Number of organizational customers of apps 5 Apps commercialized - brought to market 2 Number of app prototypes 50 Number of teams applying for in-depth support 119 Number of individual applicants to services 471 Teams that received one-on-one support 18 Table 16: Number of people trained 150 Results Number of events organized 38 Highlights for Mobile Nepal Number of developers and entrepreneurs reached 900 (mHub Nepal) Note: As reported by mLab manager. Including estimates, proxies, and underreporting due to missing data In view of the small budget and im- mobile innovation ecosystem, it has mature ecosystem, the total results of also faced major challenges. The Mobile Nepal are impressive (table 16). most important one is actually born As of March 2013, about 25 core com- out of the quick growth of the com- munity members drove activities, with munity: the developers and entre- 50 participating regularly. The commu- preneurs are hungry for more and nity organized many events, workshops, Mobile Nepal feels that it has been and some training. The eventual impact increasingly difficult to meet their of Pivot Nepal still remains to be seen, expectations. At times, community but the first indication is that it has members expressed disappointment been a most significant step towards over the hype and optimism gen- the creation of sustainable businesses. erated at events with momentum Three registered companies received quickly slowing down afterwards, as support during Pivot Nepal; one startup Mobile Nepal lacked the “bandwidth” was created as a result of the competi- to conduct substantial follow-ups tion; and four more teams are close to with teams. The organizers found it formalizing their business. Two of the hard to prioritize correctly which of finalists also applied to infoDev’s mAgri the many support demands to focus Challenge, intending to expand some of on, given limited funds available and their business into Africa. Pivot Nepal only slowly emerging support from received widespread media attention, other ecosystem stakeholders in the possibly bringing in more partners in for immature Nepalese market. As a Mobile Nepal’s effort in the future. result, Mobile Nepal was not able to give its community a clear perspec- Challenges tive and had to dampen the hopes Despite the great strides that Mobile and ambitions of some of its commu- nity members. 113 Nepal has made in building the This challenge is tied to market barriers that infoDev-funded activities provided for app startups in Nepal. Monetization to approach these units. is extremely difficult, given consum- ers’ low purchasing power (and thus Sustainability and Next Steps low potential advertising revenue) and Mobile Nepal has successfully es- the lack of payment channels such as tablished an active, self-sustaining credit cards or a mobile payment infra- community. Community champions structure. Accordingly, both large tech and new members run small sessions companies and investors do not find the on their own time and coaches and Nepalese mobile app market to be very mentors are willing to provide support attractive. Operators and the few active occasionally and informally. At the same device and operating system providers time, the community has increased its do not invest in developer outreach expectations and readiness for more. and ecosystem support programs. A There is now great demand for “hands- local angel investor community is only on” startup support. Mobile Nepal has just beginning to form, with Biruwa had to turn down many requests for the Ventures102 and the Fortune Cookie103 type of support that mLabs provide; the initiative as the most notable glimmers m2Work hackathon and Pivot Nepal of hope. These conditions have made it have only stirred more excitement and hard for Mobile Nepal to attract con- need for follow-up. crete contributions and funding from the private sector. In response, Mobile Nepal wants to continue community building, run Pivot Another challenge was Young Inno- vations’ administrative capacity in Nepal annually, and enhance this with managing a World Bank grant. Due training and acceleration programs. diligence processes and reporting re- Income sources would be diversified quirements have been burdensome for through local fundraising (mainly spon- the small startup, even after the first sorship), contract-based developer out- disbursements had been made. This reach and ecosystem-building activities has improved over time, but the lack of (for instance, hackathons and innovation institutional capacity at times caused competitions), direct contributions from delays and additional unexpected effort community members (donations, mem- for Young Innovations. bership fees), and fees from training. Finally, infoDev did not have the re- The organizers estimate that the lo- sources to build ties for Mobile Nepal cal ecosystem will still only provide to local World Bank and IFC offices limited financial support, as tech as much as was hoped for; Young partners do not tend to see Nepal as Innovations pointed out that access to a focus market and most startups are these networks provides tremendous still far from investable. As a result, value in a market like Nepal. It was only Mobile Nepal forecasts it will only be towards the later stages of the mHub able to cover a quarter to a third of its grant that Young Innovations was able expenses through income other than 114 to use the brand value and credibility additional grants. Noted startups and app products supported by Mobile Nepal (mHub Nepal) Box 10: Noted startups and app products supported by Mobile Nepal (mHub Nepal) Note: Websites at https://esewa.com.np/home; https://play.google.com/store/apps/details?id=com.nepways. nlocate&hl=en. A more comprehensive list of supported startups and teams can be found in the appendix. Detailed case studies of these and other startups will be featured in an upcoming infoDev publication, commissioned to the private sector development evaluation and research agency CAD (http://cad.globalcad.org/en/que-es-cad/). 2013 2014 2015 2016 Expenses Overhead 9,802 24,000 24,000 24,000 Training 876 4,000 4,000 4,000 Face-to-face events and outreach 3,204 15,000 15,000 15,000 Research 1,071 — — — Idea and startup competitions 6,286 25,000 25,000 25,000 Pivot Nepal 25,000 — — — Acceleration (4+ entrepreneurs) — 20,000 25,000 25,000 App testing facility — 20,000 5,000 5,000 Audit 706 — — — Total expenses 230,220 242,538 309,119 380,756 Income Service revenue and private sponsorships Contributions from Young Innovations and tech community — 4,800 4,800 4,800 Training — 1,000 2,000 2,000 Acceleration fees — — 5,000 5,000 Additional fundraising (sponsorships, etc.) — 16,000 21,000 21,000 Grants Table 17: infoDeva 40,682 — — — Budget and Financial Total income 40,682 21,800 32,800 32,800 Projections Financing gap (6,263) (86,200) (65,200) (65,200) for Mobile Nepal (mHub Total financing gap until 2016 (222,863) Nepal) Note: All amounts in US$. Budget items are kept at a high level to avoid disclosure of sensitive information. Projections are based on rough estimates, as mLabs are currently in a transition phase at the end of grant funding from infoDev. 115 a. Includes contract-based funding. 116 KEY PARTNERS KEY ACTIVITIES VALUE PROPOSITIONS CUSTOMER CUSTOMER SEGMENTS RELATIONSHIPS • MoMo co-founders (AiTi, Naiscorp), • Regular meetings & gatherings • Soft values: Exposure, awareness, • Mobile developers & tech entrepre- LaunchPad co-founders (IDG motivation • Niche / community, Topica assisting neurs locally in Hanoi, HCMC (Mobile & Beer, Mobile Weekend Capital, Open Consultant) Camps, Lunch with a Tycoon) as local event organizer • Facilitation & initiation of online • Pitching, networking, career, • Mobile developers & tech entrepre- • Ad hoc sponsors & partners for communities (Mobile Monday, business opportunities platform • Community, Topica as platform neurs across Vietnam events, workshops, & trainings (eg, LaunchPad) facilitator 5Desire, Vietnam Entrepreneurs • Peer & expert learning Foundation, Action.vn, Socbay, • Tech community leaders (bloggers, CyberAgent) • Invitations, coordination with • Niche / personal, Topica as evangelists, etc.) speakers & sponsors • Community insights dedicated networker and broker • University clubs, event location • Online courses on startup creation • Promotion platform providers • Event coordination, outreach • infoDev CHANNELS KEY RECOURCES • Events, feedback surveys • Topica’s initial network (students, investors, rural & regions) • Online communities (LaunchPad, • Topica’s incubation & PPP experience Mobile Monday) • Brand, organization capacity • Mailing lists, online communities (LaunchPad, Mobile Monday) • Speakers, expertise, business contacts • Student body The Business Model Canvas – TOPICA Mobile Social Networking (mHub Vietnam) • Staff COST STRUCTURE REVENUE STREAMS • Staff • Sponsorships & participant fees • Offline event facilitation • infoDev: mHub grant • Online marketing Note: To access the full canvas for TOPICA MSN, go to https://bmfiddle.com/f/#/PY9B8. Each relationships, channels, and value proposition elements have the according color when they mainly partner or group of partners has a specific color; activities and resources that mainly rely on a relate to the customer segment in question. Activities and resources as well as cost and revenue particular partner group have the same color. Also each customer segment has a distinct color, and streams that only relate to the infoDev grant have their own color. Topica MSN Total infoDev grant funding $35,000 Base city, country Hanoi, Vietnam infoDev grant start date March 1, 2011 First activities launched March 3, 2011 Host organization TOPICA Education Group Key partners IDG Capital, Aiti Aptech, Action.vn, Socbay, CyberAgent Targets as per grant agreement • Hosting social events, typically on a monthly Table 18: basis, to promote collaboration and innovation Basic Data in the mobile sphere for TOPICA Website www.facebook.com/groups/launchpad MSN (mHub Vietnam) Note: Funding amounts are based on actual expenses as of September 2013. Amounts are rounded to thousands. Business Model to 500 people, including ideation work- TOPICA Mobile Social Networking shops and startup showcases. Finally, (MSN) has taken mobile innovation “Lunch with a Tycoon” events gave 20 community building to scale. Through to 60 selected participants exclusive a series of events in Hanoi, Ho Chi networking access to one or two CEOs Minh City, and several provincial or high-level executives from successful capitals, combined with two vibrant companies. The organizers mentioned online communities (MoMo Vietnam that these events were popular, but also and LaunchPad104), TOPICA MSN has costly to organize. Some of the most convened a national community with promising entrepreneurs also partici- thousands of members. pated in TOPICA’s Founders Institute,105 providing at least a few with more in- Monthly events with 100 to 600 par- depth support. ticipants gave grassroots innovators the opportunity to meet successful Complementing its face-to-face ac- tech entrepreneurs, established busi- tivities, TOPICA MSN cofounded the ness people, and technical specialists. online communities MoMo Vietnam TOPICA MSN categorized the community and LaunchPad (or simply “Launch”). in participant tiers and ran three series They have become the virtual go-to of events: “Mobile & Beer” can be de- places for everyone interested in the Figure 20: scribed as a typical Mobile Monday event Vietnamese mobile app and startup (opposite page) and is the most informal and inexpensive founder spheres. LaunchPad reaches The Business event form. The TOPICA team described beyond “mobile” and is used by many Model Canvas – Mobile & Beer as the most effective and early-stage entrepreneurs to pitch TOPICA Mobile efficient form of face-to-face community ideas, search for cofounders, and Social Networking building. “Mobile Weekend Camps” were contact role models and mentors—not (mHub Vietnam) organized as one-day bootcamps for up least because of active facilitation by 117 Indicator Result as of June 2013 Consumer customer traction: Number of app downloads /users/subscribers 100,000 Organizational customer traction: Number of organizational customers of apps 15 Apps brought to market 60 Number of individual applicants to services 2600 Table 19: Teams that received in-depth support 10 Results Number of people trained 100 highlights for TOPICA Number of events (additional virtual events) 46 (4) MSN (mHub Number of developers and entrepreneurs reached 9,000 Vietnam) Note: As reported by TOPICA MSN team. Including estimates, proxies, and underreporting due to missing data. TOPICA complemented Mobile Monday events with other activities such as its Founders Institute program or its “cofounder dating” sessions. It is hard to discern the results of TOPICA MSN from the overall numbers that TOPICA has observed. IDG Ventures,106 an ICT-focused ear- networking sessions that would appeal ly-stage investment fund. The LaunchPad to larger groups of people. So, TOPICA platform has also become a space for MSN has contributed to the setup of a community leaders to seek feedback national innovation and startup com- and opinions from the community, and munity where there was none before. to share market information and trends. Both communities have taken on the Positive network effects due to the function of informal job boards. size of the community and exponential growth of startups that came out of the Lastly, TOPICA MSN introduced an on- activities led to impressive results for line course with guidance how to start a an mHub (table 19). An estimated 500 company in Vietnam. TOPICA MSN de- jobs at startups were created; this in- veloped the curriculum based on a sur- cludes both jobs added at startups that vey with about 1,000 business profes- had existed before and jobs at about 20 sionals. Later, the course reached about newly founded startups. 20,000 students across the country after As another positive result, TOPICA MSN it was handed to the Vietnamese Young increasingly attracted stakeholders from Entrepreneurs Foundation for rollout. groups beyond the core developer and entrepreneur community. For instance, Results in March 2012, Vietnam’s prime minister TOPICA MSN has succeeded in build- and other government officials participat- ing two large mobile innovation online ed in an event in Hanoi. New government communities that span the country, and initiatives, for instance, by the National also engaged developers and entrepre- Agency of Technology, Entrepreneurship neurs offline through a great number and Commercialization (NATEC) aim to of events. Interviewees said that small learn from and possibly collaborate with pockets of developer communities had TOPICA. TOPICA has also learned from existed before the TOPICA MSN project, the MSN project to rollout and scale an but that there was no national commu- infoDev-supported Virtual Incubation 118 nity and never any momentum to run program107 for provincial capitals. Challenges turned into mere “talk shows.” They TOPICA’s MSN project encountered a acknowledged that it was hard to cater range of challenges. First and foremost, for both participants that had been the organizers pointed out arduous active for months and years and those reporting and grant administration re- that had just joined the community. The quirements. Even though TOPICA had clients suggested that events be more worked with the World Bank before, varied, specific, hands-on, participa- the small MSN team was overwhelmed tory, and involve fewer participants—a with the steps mandated by World Bank request that TOPICA MSN is aware of grant procedures. The team felt that the and aims to address (see Sustainability small grant did not merit such intricate and Next Steps). Other suggestions administration, and that infoDev could for improvements included the es- have been more “hands-on” to limit tablishment of a facilitated open code the effort required by TOPICA for grant repository for MoMo and LaunchPad management . members, and a stronger engagement of FPT University that was deemed to Also related to the small grant size is add technical skill and many talented the missed opportunity of thorough coders to the community. monitoring and evaluation. The TOPICA MSN team would have liked to conduct Sustainability and Next Steps a more sophisticated analysis of the The grant for TOPICA MSN ended in widespread impact that the grant has August 2012. Since then, the num- had, but felt this was impossible with ber of subscribers for the two online “I can’t the resources provided. It became clear that a community and ecosystem communities has grown further, with imagine LaunchPad now counting about 10,000 focused initiative such as TOPICA MSN members. Mobile & Beer events could a day likely has far-flung positive impact, but be continued with sponsorship, in-kind that elaborate assessment methods contributions by organizers, and small without would be cost-prohibitive. Again, the financial contributions from TOPICA Launch.” team pointed out that infoDev could take and other core partners. However, the on monitoring and evaluation tasks and missing core budget previously avail- A blogger suggested that better and more effi- able from the grant meant that organi- from Hanoi cient methodologies to assess indirect zation and fundraising have to be done positive effects of community building “on the side,” increasing the lead time initiatives be developed. to about three months, so that events can only be run at quarterly intervals. As far as TOPICA MSN’s operations are concerned, the team found that Next, TOPICA wants to expand the incumbent technology companies were MSN project across the region, part- sometimes unwilling to contribute to nering with existing Mobile Monday the startup community, as they felt chapters and tech incubators. The they would nurture threats to their own TOPICA MSN team also wants to re- business. As a result, at times, it was spond to requests from the communi- hard to crowd together “big guys” and ty to use more hands-on, interactive, “small guys.” and focused event setups. In partic- The interviewed entrepreneurs and ular, the organizers want to facilitate event participants mentioned that the subcommunities, for instance, on the setup of Mobile & Beer events had topic of mobile games. 119 become overused and that they often 120 KEY PARTNERS KEY ACTIVITIES VALUE PROPOSITIONS CUSTOMER CUSTOMER SEGMENTS RELATIONSHIPS • Ad hoc contributors (individual • Mobile Garage (community building • Employability, entrepreneurship • Female high school graduates & entrepreneurs, tech executives), & technical skills • Niche, Akirachix teachers & mentors university students interested in universities, bootcamps, girls volunteers hackathons, app showcases) address student groups in tech • Building confidence & motivation • High school outreach • Awareness raising in areas outside • Trainers • Niche, Akirachix as platform • Idea stage app developers • Ad hoc events (MoMos, Girls in of Nairobi facilitator ICT Day) • Location & outreach partners • Women entrepreneur & tech • Pitching & networking platform, (SODNET, iHub, universities) • Community, Akirachix as community community at large • Training (ICT skills, graphic design, competitive environment leaders soft skills) • Funding, sponsorship, & in-kind • Establishment of a women tech donations (infoDev, Google RISE, • Hosting events & trainings, sub-community Computer Aid, Seneca Group) online & offline outreach • Motivation & guidance to start chapters outside of Nairobi • Event organization, coordination, CHANNELS admin, outreach & blogging, etc. • Teachers in courses, mentors The Business Model Canvas – Akirachix (mHub Kenya) at events KEY RECOURCES • “Multipliers,” local community facilitators • Mentors’ expertise on entrepreneur- ship & tech, volunteers’ passion • Event outreach & communication • Trainers’ experience & motivation • Social media, website, online • Space & outreach channels channels • Personal interaction in iHub space • Akirachix staff time, equipment, food & swag for events COST STRUCTURE REVENUE STREAMS • Administrator • infoDev: mHub grant • SIDA contribution • Trainers • Google Rise sponsorship • Small sponsorships for events & competitions Note: To access the full canvas for Akirachix, go to https://bmfiddle.com/f/#/N3G53. Each partner relationships, channels, and value proposition elements have the according color when they mainly or group of partners has a specific color; activities and resources that mainly rely on a particular relate to the customer segment in question. Activities and resources as well as cost and revenue partner group have the same color. Also each customer segment has a distinct color, and streams that only relate to the infoDev grant have their own color. Women-targeted, grassroots tech community building and training Akirachix Total infoDev grant funding $53,000 Funding for technical assistance $41,000 (approximate) Base city, country Nairobi, Kenya infoDev grant start date May 1, 2011 First activities launched September 10, 2011 Host organization Akirachix Key partners iHub Nairobi, mLab East Africa, Google Kenya, Samsung Kenya, Swedish International Development Agency, University of Nairobi, Jomo Kenyatta University of Agriculture and Technology, Egerton University, African Nazarene University, Maseno University, Masinde Muliro University of Science and Technology Targets as per grant agreement • Hosting social events, typically on a monthly basis, to promote collaboration and innovation in the mobile sphere Website akirachix.com/ Table 20: Social media and other web resources @Akirachix, https://www.facebook.com/pages/ Basic Data AkiraChix/173842546011090, for Akirachix http://akirachix.com/Blog/ (mHub Kenya) Note: Funding amounts are based on actual expenses as of September 2013. Amounts are rounded to thousands. Business Model through bootcamps, hackathons, and Akirachix has become a leading com- app showcases. Over a few days, par- munity and capacity builder for female ticipants receive brief training on tech tech entrepreneurs in Kenya.108 The topics, are introduced to mentors and focus is less on startup creation and Akirachix community leaders, and more on basic skills development, sometimes participate in small com- empowerment, and community build- petitions. The high school outreach ing. To this end, Akirachix ran the program aims to open up a perspective Mobile Garage program, high school in the tech industry for graduates that outreach, training, and various ad hoc are unsure about their future. Also the events and activities.109 training program targets high school graduates, teaching them computer Figure 21: Akirachix’s main customer segments literacy, graphic design, programming, (opposite page) comprise young, talented women that and soft skills over the course of one The Business have an interest but no prior knowl- year. Finally, additional events and Model Canvas – edge or experience in ICT. The Mobile community building through online Akirachix (mHub Garage program reaches out to stu- and offline outreach and communica- Kenya) 121 dents at universities across Kenya tion are helping to build a larger tech entrepreneurship community for women team of five women entrepreneurs and in Kenya, beyond any specific project. community leaders, tens of volunteers contribute to organizing events, blog- The added value is three-fold. First, ging, graphic design, outreach, and so girls’ and young women’s skills are forth. Larger networking events, such enhanced, starting from a basic level. as the Girls in ICT Day,110 help to reach Bootcamps spark interest and give a a broader group of stakeholders and glimpse of potential professional path- engage new volunteers. ways. Akirachix has designed an inter- disciplinary training curriculum and has Mentors and volunteers contribute secured equipment and space. These simply for personal satisfaction, with- programs aim at higher employability out material compensation. Teachers based on technical skills, as well as receive some pay, but this is done improved entrepreneurship and busi- more for reasons of accountability ness skills. Second, intangibles such and commitment than to compensate as confidence and aspiration are just as them for the substantial time and important. Role models and mentors effort that they invest. Akirachix has play a crucial role in highlighting a path good staff relations and mission fit to professional success for client girls with the iHub, which takes in a special and women. App showcases address partnership role: Akirachix makes the slightly more advanced innovators extensive use of the iHub space for and instill a culture of competition and events and team meetings, recruits performance for university students (in trainers and mentors from the iHub addition to a platform for exposure and community, and is featured frequently networking). Third, Akirachix makes a in iHub media and outreach. Google conscious effort to reach groups that RISE became a major sponsor111 in otherwise would not receive such val- 2012, and several organizations (such as SODNET, Computer Aid, Seneca ue: not only does the programming aim Group) helped with in-kind contribu- to bridge the gender gap in ICT, it also tions and free space. targets universities in cities outside of Nairobi (such as Kisumu, Kakamega, Results and Mombasa) and schools in impover- ished areas of Nairobi. Akirachix has established a powerful and credible brand in African wom- All projects are embedded in the larg- en tech entrepreneurship and built er effort to establish the women tech a strong and motivated community. community as a strong subcommunity Even though it was hard for the grass- in the Kenyan tech entrepreneurship roots organization to manage the in- ecosystem. Akirachix employs an “or- foDev grant, the learning process also ganic” community building strategy, helped it mature as an organization based on dedicated community cham- and to establish administrative struc- pions and word of mouth (both online tures and capacity. This, in turn, was and offline), fueled by occasional media instrumental for securing follow-on 122 mentions. Beyond the Akirachix core funding from SIDA.112 Indicator Result as of June 2013 Women that found employment in tech as a result of support 23 Number of app prototypes 58 Number of individual applicants to services 491 Number of teams that received in-depth support 21 Women-led teams that received in-depth support 12 Number of people trained 396 Number of people trained intensively (one-year course) 36 Number of events organized 12 Table 21: Number of developers and entrepreneurs reached (estimated Results for Akirachix share of women) 1,992 (70%) (mHub Kenya) Note: As reported by Akirachix manager. Including estimates, proxies, and underreporting due to missing data. Brand and community related out- the small size of the grant bore no comes are hard to pinpoint, but several relation to the bureaucratic burden indicators provide a glimpse. Almost that came with it and, unexpectedly, 400 participants received brief training supposed project implementers had at three-day bootcamps across Kenya. to take on roles as grant administra- Thirty-six girls graduated from the tors. Akirachix conceded that it had in-depth, one-year training program, underestimated the requirements and which led to 23 of them successfully necessary resources, but it felt that applying for jobs. Akirachix has more infoDev could have alleviated the bur- than 200 community members and es- den by recruiting a local consultant timates that it has reached nearly 1,400 who could have supported auditing women developers and entrepreneurs as well as monitoring and evaluation. in total. Anecdotal evidence shows that Similarly, the positive ripple effects local subcommunities are beginning to of Akirachix’s projects are probably form: for instance, students founded many, but they are often indirect and a tech meetup series in Kisumu. The may only become apparent later. Akirachix story also made it into U.S. Although Akirachix would have liked media through a feature on NPR113 and to conduct more analysis through honorable mentions in many press follow-ups with clients on its own, re- articles such as an op-ed by Finnish sources were simply insufficient to do Minister Heidi Hautala.114 this. The team suggested that infoDev explore contractual agreements in- Challenges stead of grants for grassroots orga- The major challenges that Akirachix nizations, as this could give greater faced were related to the limited ad- freedom to specify customized deliv- ministrative resources of a grassroots erables and reporting duties, and set a organization. The team found that clearer framework for implementation. 123 From implementation experience, participants should be required to pay a Akirachix found that, despite general token amount to ensure greater com- excitement and passion for its proj- mitment. Akirachix also introduced a ects, achieving meaningful and lasting small salary for teachers. This was not impact was often challenging. Girls intended to fully compensate teachers, and young women with no prior work but rather to tackle the occasionally or ICT experience needed continuous lacking reliability of volunteer teachers mentoring and commitment from that had occurred at the beginning of teachers. Often, social issues, gender the program. Several interviewees em- stereotyping, and life choices got in phasized that Akirachix should expand the way of clients’ pursuit of a career its current training and grow revenues in tech. Mentors and teachers also from sponsorship. realized that, while it is important to be supportive, there also needs to be However, Akirachix also found that milestones and consequences if clients the search for sponsorship can have and students do not keep promises. its downside. The team thought that some of the prospective sponsors for In particular, Akirachix learned that par- bootcamps within the Mobile Garage ticipation fees have ambiguous effects: program would push their own agen- if the goal was to achieve the broad- da or look for concrete returns, while est possible engagement and reach, failing to endorse the grassroots charging participants would be coun- principles that Akirachix embraces. terproductive, and instead Akirachix Some interviewees suggested that at- asked for in-kind contributions and vol- tracting sponsorship might be difficult unteering in return. Interviewees point- for Akirachix because it would mean ed out that Akirachix ought to formalize a subtle loss of focus and shift away this process and recruit volunteer “am- from the core mission of bringing more bassadors.” In turn, some interviewees women into tech entrepreneurship. 124 expressed that, at least for the training, These stakeholders pointed to the example of open competitive events such as App Showcases, which feature only a few female participants, if any. Several interviewees also pointed out “Akirachix’s main that Akirachix could still improve its customer segments organizational rigor, for example, by consist of young, formalizing mentorship recruitment and partnership building, or by being talented women that more assertive and confident regard- have an interest but ing the level of decision makers that no prior knowledge Akirachix reaches out to. Other ideas for improvement include the establishment or experience in ICT. of Akirachix chapters at universities across the country, or an internship program that would take clients out of the now-comfortable iHub world into revenues, investments, and jobs. In the professional environment of tech line with this strategy, the organiza- startups and corporations. tion is in the process of receiving a Akirachix also found it hard to partner grant from SIDA115 that emphasizes directly with mLab East Africa, which the enhancement of access to and had been an original goal for colocating use of ICT for women and girls. As a an mHub and an mLab. From Akirachix’ long-term vision, Akirachix would like perspective, the mLab did not seem to to formalize a women innovation hub have the capacity or determination to with its own physical space to ensure establish a privileged role for Akirachix. greater continuity. The focus on women tech community Over the next four years, Akirachix building appeared secondary to the aims to diversify its revenue streams mLab’s areas of focus. and generate income from consulting services and training participation Sustainability and Next Steps fees, sponsorship (for example, from Building on its success and popularity, Samsung Kenya to sponsor part of Akirachix aims to continue and further the annual Mobile Garage Exhibition), expand its current activities. Akirachix grants from government and donors. It sees the greatest need for skills devel- also wants to extend its partnerships, opment, empowerment, and opening for instance, with NaiLab to deliver up other opportunities for women and activities jointly. Akirachix forecasts girls with little or no entrepreneurial growth of about 30 to 35% per year. Its and professional experience. This major expenses are for salaries, trans- also means that Akirachix will further port, and equipment. It estimates that de-emphasize infoDev’s and mLabs’ it will require about $15,000 per year in core goals that focus more on direct additional funding to sustain activities 125 startup creation and the creation of in the next few years. At the heart of Uganda’s mobile Mobile Monday Kampala innovation ecosystem (mHub Uganda) Total infoDev grant funding $35,000 Funding for technical assistance $41,000 (approximate) Base city, country Kampala, Uganda infoDev grant start date February 25, 2011 Host organization Mobile Monday Kampala Key partners Hive Colab, Grameen Foundation AppLab, @The HUB Kampala, Mara LaunchPad, FinAfrica, Outbox Hub, SMS Media, Orange Uganda, Smile Communications, Yo, Send Airtime, Google Uganda, UGO, Text To Change, D-Mark Mobile, Mountbatten, Owino Solutions, True African Website http://www.momokla.ug/ Table 22: Basic Social media and other web resources @MoMoKampala, Data for Mobile https://www.facebook.com/pages/ Monday (MoMo) Mobile-Monday-Kampala/145954058768730, Kampala (mHub http://www.mobilemonday.net/category/kampala Uganda) Note: Funding amounts are based on actual expenses as of September 2013. Amounts are rounded to thousands. Background and Achievements mentors, and other stakeholders. The MoMo Kampala chapter was found- Discussions revolve around the needs ed in January 2010. The project leader of the tech community and educating started it as a side project without ded- clients on startup methodologies and icating substantial time and effort to it. trends in the mobile space. In this way, This changed when the mHub grant was the MoMo events also became gath- awarded by infoDev in 2011. While the erings for subcommunities in mobile project leader would still only receive health, mobile agriculture, ICT for a small salary, financial compensation development and other topics. In total, freed him up to work more consistently MoMo Kampala estimates that it has and continuously on the project and reached about 2,000 developers and build the local MoMo brand. entrepreneurs. Over the two years of grant duration, A Truly Grassroots-Driven, MoMo Kampala grew substantially. Collaborative Approach One event was hosted every five to six The MoMo Kampala leader has deliber- weeks, with participation ranging from ately maintained a grassroots-oriented about 100 to more than 200 people. approach. The MoMo chapter con- From February 2012 to August 2013, siders developers and entrepreneurs MoMo hosted or was affiliated with a to- as the lifeblood of the ecosystem as tal of 38 events. Individuals and teams a whole. MoMo Kampala has never 126 of developers and entrepreneurs get sought to occupy a set physical space introduced to investors, tech partners, but rather utilizes a series of events at varying locations in order to span mul- base. MoMo has also partnered with tiple stakeholder communities. These other local innovation hubs such as gatherings almost always emerge spon- the Grameen Foundation AppLab,117 @The taneously out of informal discussions HUB Kampala,118 the Mara LaunchPad,119 and previous events, and are organized FinAfrica,120 and most recently the Outbox within time frames of about four weeks Hub.121 Under the Business Innovation or less. It is important for all organizers Consortium Uganda (BICU),122 the MoMo to “keep an ear to the ground” and pick Kampala lead has made an effort to up new trends and interesting topics as create a vehicle for more systematic they emerge. The MoMo leader swiftly and continuous collaboration and ex- establishes an organizing team for each change between the innovation hubs, event, drawing team members from his but the new informal organization has personal network of local industry and yet to find a clear mandate and gover- community leaders, depending on the nance model. given trend or theme that the event will MoMo Kampala has deliberately re- address. Whenever other team leaders fused to structure and formalize its are willing and able to take the lead on operations more, or to make long-term the event organization (for instance, if commitments. The flexible, nonhierar- they have access to the resources of chical, and informal modus operandi one of Kampala’s innovation hubs), the is, it believes, instrumental to address MoMo lead will limit himself to a sup- the needs of the developer community porting and brokerage role. and the ecosystem as a whole. The MoMo Kampala has become the nexus MoMo chapter has also made it a goal for several other emerging activities to directly point out shortcomings in in the mobile innovation ecosystem, the ecosystem and to challenge in- in particular several innovation hubs cumbents and large organizations that in Kampala. The MoMo chapter grew find it hard to embrace and support a “organically” and partnered with dif- community-oriented approach. This ferent hubs and other organizations in does not mean that MoMo Kampala the ecosystem wherever it felt that the aims to exclude participation and con- collaboration would benefit the ecosys- tribution from large organizations, just tem and address an existing need. The that this needs to happen on its own Hive Colab116 innovation hub takes on a terms. High-level executives from tech special role, for one, because the MoMo companies, policy makers, and other Kampala leader is also on its board of role models and decision makers have directors, but also because it is an open been welcome guests and speakers for developer and entrepreneur cocreation MoMo events, and they often function space, partly modeled on Nairobi’s as informal mentors for MoMo’s clients. iHub. Many MoMo events were hosted in the Hive Colab location, contributing Fundraising and Sustainability to the buzz in the space and leveraging In line with its flexible, grassroots-ori- its community of resident developers ented approach, MoMo Kampala oper- 127 and entrepreneurs as a participant ates on a shoestring budget and does capacity now that infoDev support has ended. It is a deliberate strategy for MoMo Kampala to accumulate small amounts of sponsorship funds, so as to “[MoMo Kampala] is not create a dependence on a single or eminently replicable. few partners, and to maintain MoMo’s It is just about looking inclusive, participatory approach. around and seeing Accordingly, MoMo Kampala believes what needs to be done that it can be an illustration for the pow- er of the MoMo approach that is based for the ecosystem. on a grassroots and community ideal. Then go ahead and The leader encourages others to see do it, and you will get MoMo Kampala as replicable, and he has started to engage with stakehold- the encouragement ers in Rwanda and Namibia on how to to continue.” expand their local ecosystems through a community-based approach. Quote from interview with Daniel Stern, leader of MoMo Kampala Challenges MoMo Kampala’s work has not been without challenges. In particular, the strong emphasis on grassroots en- not necessarily aim to expand into gagement and impartiality has made a much larger project. Sponsorship it harder to attract large sponsorship; comes in small amounts of about $400 often, for corporations, specifying a per sponsoring organization every six long-term agenda and concrete de- months. Such small sponsorships and liverables was a prerequisite. MoMo in-kind support came from SMS Media, Kampala is also hard-pressed to find Smile Communications, Yo, Send ways of assessing its positive impact Airtime, UGO, Text To Change, D-Mark and document its relevance and value Mobile, Mountbatten, Owino Solutions, addition more concretely. The effects of and True African. Slightly larger contri- MoMo Kampala’s community building butions of more than $1,000 came from are notoriously difficult to track, as Google Uganda, with a similar contri- they are indirect and only materialize bution expected from Orange Uganda. through other ecosystem stakeholders The MoMo chapter leader describes the later making concrete decisions to infoDev grant as seed funding that em- partner, invest, support, and so forth. powered him to build up the brand and support the ecosystem independently. At the same time, MoMo Kampala has Through many small sponsorships, realized that more continuity in develop- MoMo Kampala has become sus- er and entrepreneur support is needed tainable, in the sense that events are than the ecosystem is currently offer- 128 still being hosted at about the same ing. MoMo Kampala acknowledges that startup support still has wide gaps, and that the available talent cannot fully un- fold its potential with the help of MoMo and fledgling innovation hubs alone. At the same time, MoMo Kampala has neither the means nor the mission to provide the support mechanisms that are urgently needed, such as access to finance, incubation and acceleration, structured mentorship brokering, poli- cy reforms. It does expect, however, to facilitate a more structured collabora- tion between Uganda’s innovation hubs, which could lead to more efficient use of existing resources and to attracting new ones. 129 Endnotes 51 http://www.businessmodelgeneration.com/canvas. 52 See, for instance, http://businessmodelhub.com/page/business-model-canvas or http://www.youtube.com/ watch?v=QoAOzMTLP5s. 53 http://www.businessmodelgeneration.com/book. 54 A helpful discussion on the difficulties of categorizing stakeholders as partners or customer segments can be found at http://businessmodelhub.com/forum/topics/confusion-in-business-model-canvas-customers-or-partners. Moreover, a more analytically sound method to include donors and impact investors as distinct types of customer segments is the double-sided Business Model Canvas for NGOs (see http://www.innovativenonprofit.com/2012/08/the-nonprofit-business-model-can- vas-2/#.UmG_Hvn3Hn8). This report did not use this method to maintain simplicity. 55 http://www.ihub.co.ke/. 56 The competition’s first edition in 2011 was dubbed “Pivot 25” but, for the sake of simplicity, this report will refer to all three editions in the series as “Pivot East.” 57 The first waves of trainees were trained for six months before the schedule was tightened and aligned to university calendars. 58 http://www.ihub.co.ke/blog/2013/06/mlab-east-africa-outcomes-two-years-later/. 59 The coaches had at least monthly meetings with startups to discuss their business and fundraising strategy and how startups could conduct and measure market experiments most effectively. A few incubatees continued to contract Viktoria Solutions at their own expense once mLab’s subsidy ended. Additionally, the startups also received in-kind mentoring services from other mentors that were brokered by the mLab. 60 See, for instance, http://techcrunch.com/2013/09/25/african-accelerators-change-tack-as-savannah-valley- emerges-from-the-wilderness/. 61 Starting in October 2013, the mLab is running a virtual incubation program for startups spread across the region, recruited from among the 25 finalists of Pivot East 2013. The project is funded through a $30,000 contribution from AfriLabs, as a tranche from an infoDev grant. http://www.infodev.org/articles/east-african-virtual-incubation-pilot-launch-nairobi-kenya. 62 http://www.theinnovationhub.com/. 63 http://gic.theinnovationhub.com/. 64 http://www.csir.co.za/meraka/. 65 Nokia also played an important role in the setup phase of the mLab, including brokering linkages to DST. 66 This implies that DST is not represented as a customer segment in the mLab’s business model canvas. For a more detailed explanation, see the section on Applying the Business Model Canvas to mLabs and mHubs. 67 http://www.windowsphone.com/en-us/store/app/mlab-studio/962078da-52a2-4598-9704-5b25ef6485cd. 68 http://www.ungana-afrika.org/. 69 http://www.saisprogramme.com/. 70 http://www.mct.gov.mz/portal/page?_pageid=615,2812864&_dad=portal&_schema=PORTAL. 71 http://www.tia.org.za/. 72 http://www.saine.co.za/. 73 http://www.raizcorp.com/entrepreneurs/partner-elite/partner-elite-ramp/. 74 http://www.bandwidthbarn.org/. 75 http://www.raizcorp.co.za/ . 76 http://www.balefyre.co.za/products-ussdx.html. 77 http://nacloud.dimensiondata.com/Cloud-Services/Overview%20. 78 http://www.techtransfer.csir.co.za/2012/09/mobi4d-mobile-service-delivery-platform/. 79 mLab ECA serves entrepreneurs in Eastern Europe, South Caucasus, and Central Asia. 80 Enterprise Development and Market Competitiveness (EDMC) is a USAID initiative to spur local economic growth: http:// www.edmc.am/. 81 In particular, CRDF ran a “From Mind to Market” grant competition in 2012 within its Science & Technology Entrepreneurship Program (STEP): http://www.mlabeca.com/eng/news-announcements/ quot_from-idea-to-market_quot_-grant-competition_-the-deadline-of-submission-of-application-forms-is-prolonged/. 82 http://www.gov.am/en/news/item/6755/. 130 83 http://www.micarmenia.am/en. 84 http://www.armindia.am/. 85 http://gitc.am/gitc/. 86 http://infodev.org/press-releases/media-release-5th-global-forum-innovation-and-technology-entrepreneur- ship-awards-five. 87 http://www.mlabeca.com/eng/news-announcements/mlab-team-won-the-first-place-at-imagine-cup-armenia-2012/. 88 http://digitec.am/files/competition_ml/agenda_digitec_business_forum.pdf. 89 http://armenia.startupcup.com/. 90 http://www.t2vc.com/armenia/. 91 http://www.techinasia.com/5-top-winners-mlab-east-asias-mobile-innovation-challenge/. 92 Although prominent voices in the discussion on startup and innovation ecosystems warn of heavy-handed government intervention. See, for instance, Lerner, 2012; or a recent discussion at http://www.technologyreview.com/news/516501/ in-innovation-quest-regions-seek-critical-mass/. 93 http://www.younginnovations.com.np/. 94 http://www.mobilenepal.net/events/forum-mobile-ecosystem-nepal-exploring-opportunities-and-challenges. 95 http://m2workhackathon.org/. 96 http://m2workhackathon.org/sites/default/files/downloads/Report_m2Work_Hackathon.pdf 97 http://pivot.mobilenepal.net. 98 http://www.nyef.org.np/. 99 http://opennepal.net/. 100 http://www.developmentcheck.org. 101 http://www.hackathon2013.vawhack.org/. 102 http://www.biruwa.net/. 103 http://nepalitimes.com/news.php?id=13087#.UkVDRoauwYs. 104 https://www.facebook.com/groups/launchpad. 105 http://e27.co/2013/06/19/supporting-vietnams-tech-startup-growth-from-the-ground-up/. 106 http://idgvv.com.vn/en/. 107 http://www.globalforum2013.co.za/downloads/presentation/Day%202/Stream%20C%20-%20Incubation%20 LabSession%203/Dang%20My%20Chau%20Presentation.pptx 108 Akirachix recently posted a summary of their projects, available at http://akirachix.com/Blog/our-3-year-journey-part-12/ 109 The mHub grant funds that Akirachix received from infoDev served mainly to fund the Mobile Garage project, hire an administrator, and purchase computer equipment for training. Other Akirachix activities were strengthened indirectly. 110 http://akirachix.com/Blog/akirachix-to-host-girls-in-ict-day-on-27th-april-2013/. 111 http://akirachix.com/Blog/akirachix-wins-google-rise-award/. 112 http://www.sida.se/English/current-topics-archive/2013/One-woman--one-mobile-phone/. 113 http://www.npr.org/blogs/alltechconsidered/2012/12/24/167961947/kenyan-women-create-their-own-geek-cuture. 114 http://www.project-syndicate.org/commentary/women-on-the-verge-of-an-economic-breakthrough. 115 http://www.sida.se/English/current-topics-archive/2013/One-woman--one-mobile-phone/. 116 http://hivecolab.org/. 117 http://www.grameenfoundation.applab.org/team/uganda.html. 118 http://thehubkampala.com/. 119 https://www.facebook.com/MaraLaunchpad. 120 http://finafrica.org/. 121 http://www.outbox.co.ug/. 122 http://ict4entrepreneurship.com/2012/07/26/meet-the-business-innovation-consortium-uganda-bicu/. 123 http://www.universalia.com/en/services/monitoring-evaluation. 131 Appendixes Appendix A. Aggregate mLab and mHub Results Indicator Result as of June 2013 Revenues generated by startups $1,119,285 External investments raised by startups $2,605,905 Number of startups/entrepreneurs that secured external investments 39 Direct jobs created by startups 283 Other jobs created or brokered 662 Number of jobs created for women 162 Other jobs created or brokered for women 129 Startups created 97 Apps monetized 218 B2C customer traction: Number of app downloads/users/subscribers 2,661,600 B2B/P customer traction – Number of enterprise/public agency customers, or customers that contract development of the app 5,706 Number of notable social development apps broad to market (targeting pro-poor, base of the pyramid, mHealth, mAgriculture, rural populations, etc.) 102 Solid startup teams 101 Apps brought to market 516 Number of app prototypes 782 Number of teams applying for in-depth support through mLab/mHub 1,180 Number of individual applicants 5,624 Teams that received in-depth, one-on-one support (mentoring, coaching, etc.) 292 Women-led teams that received in-depth, one-on-one support (mentoring, coaching, etc.) 38 Number of startups/entrepreneurs that secured investments from mHub/mLab 51 Number of people trained (any training, including daylong workshops) 2,535 Number of people trained intensively 315 Number of physical, face-to-face events organized 376 Table A.1: Complete Number of virtual events 25 mLab and Number of developers and entrepreneurs reached 31,981 mHub Results Number of women developers and entrepreneurs reached 4,967 data Note: As reported by mLab and mHub managers and local consultant for mLab East Asia. Including estimates, proxies, and underreporting due to missing data. Including results for mHubs in Azerbaijan, Georgia, Moldova, and Tanzania, which were not covered as case studies. 132 Appendix B. List of mLab and mHub Affiliated Startups and Startup Teams mLab/mHub Startup/team name App name mLab East Africa Zege Technologies mPayer Pluspeople Kenya LTD Uhasibu Space Kenya Networks Ltd Whive MTL Systems Limited MFarm Limited MFarm Eneza Education Eneza Education Shimba Technologies MedAfrica Kopo Kopo Kopo Kopo mLab Southern Africa Afroes Haki, Moraba Sowertech AftaRobot/Tour2.0 Jatamobile African Mobile Awards/Tuneme Innoprenez CareerWiki CytoTouch Cyto Digisense Bchat PassReview PassReview BookBay Bookbay Geekulcha AppChemy GoMetro (app development), TravelWiki, Takealot, others Mfactory GoMetro (app development), PassReview, Student 101, eReader, others Quirio Quirio Student 101 Student 101 (app development) Realapp Realapp GoMetro GoMetro/GoGauteng Limitless Mobile Multiple bSmart bSmart Music of Africa Zizo/MOA mLab ECA NaKo Games Princess Cakes, Fashion Girl, Hello Kitty, The Gardener MySales Paint In, MySales MicroForester MicroForester SmartStream SmartStream Acting Software Stories HUB Team led by Narek Clean H2O, Paint the Music, App review, Hayrapetyan 4Seasons shaker, Anime logo quiz Anorayr Yerevan ATM mTech SOS, WinPhone Games, other enterprise apps iMixer iMixPlayer moGoni PoliceCameras Gyumri teams microPictureTranslator 133 Appendixes Appendix B. List of mLab and mHub Affiliated Startups and Startup Teams (continued) mLab/mHub Startup/team name App name mLab East Asia Lifebox Lifebox Daily New Way Street Foods Guide Red Team Fashion House Seahorse KID’S GARDEN iSmartbike iSmartbike Watermelon Bach Dang Battlefields Thasa Anti-Counterfeiting App Tapi Tapi Linpax Linpax Gao Group Stick Ninja MUT-UTT MGTS Taxi mHub Nepal nLocate n Locate eSewa eSewa Birthday Forest Birthday Forest Mystic Vision M-Varnamala StoresMunk StoresMunk Mobile Survey Mobile Survey BlametheStars.com BlametheStars.com mHub Vietnam Wala Wala Appota Appota Tapmee Tapmee Delta Viet Tiki.vn Tiki Yton Blue Up Money Lover Money Lover Kleii Kleii mHub Kenya MFarm Limited MFarm 254 Events 254 Events Mbegu Halisi Mbegu Halisi Table B.1: Helpful verses Helpful verses List of mLab JMavuno JMavuno and mHub Bonyeza, Bonyeza, Affiliated Kemobi Kemobi Startups Note: As reported by mLab and mHub managers. Startups/teams included might have ceased operation. Type and intensity of support that startups/ teams benefited from varies. Not all startups/teams have been created as a result of mLab or mHub activity. 134 Appendix C. Collection and Analysis of Results (Radar Charts) All data were collected through self-reporting and interviews with mLab and mHub managers (and in mLab East Asia’s case, the local infoDev consultant). All data need to be seen in the context of attributes of local ecosystems and the challenges and opportunities that any given mLab or mHub faced (see section Why Evaluate Business Models of mLabs and mHubs?). Several—if not most—of the measured indicators are difficult to standardize and track reliably. This report makes the case that mLabs and mHubs are embedded in complex startup and innovation ecosystems, and that their holistic impact is notoriously difficult to quantify. Attribution of impact is a key problem in startup ecosys- tems where actors are intertwined in various and complex ways and effects often take time to become apparent. Hence, the results should only be seen as indicative, until feasible, more sophisticated evaluation and impact assessment methods become available for ecosystem-oriented interventions such as innovation hubs. The percentage values in the radar charts in the section on Results for mLabs and mHubs, as well as in the case studies represent values standardized across data for the four mLabs, with the mLab that had the highest result receiving 100%. Several indicators are composed of weighted subindicators. The weights were determined based on the relative relevance of the subindicators towards the goal of sustainable startup creation (see table C.1). Indicator Subindicator Definition (weight) Revenue generated Includes funds received by supported startups from their by startups customers in exchange for the service or product that they are offering. This does not iinclude grants and in-kind contributions to startups. Investments Includes equity investments, convertible debt and similar raised by startups instruments, grants, and other financial contributions raised by the startups. This does not include investments made by the mHub/mLab itself. Jobs created Includes jobs that were created by the startups or solid and supported startup teams which were supported. Includes founders. Startups Startups (2/3) Only includes formally registered, incorporated companies. and startup teams created Solid startup Refers to teams that have formed, are in the process of setting teams (1/3) up a startup, and that are likely to incorporate in the near future. App customer Consumers (2/3) Pertains to “business to consumer” apps. Downloads/ traction subscribers/user numbers etc. are used. Organizations (1/3) Pertains to enterprise apps, apps that target public agencies as customers, or other app business models that rely on few large customers. App innovation Apps Includes apps that have generated some money (sold for a fee to a consumer, (apps monetized, monetized (1/2) made advertising revenue, sold to an enterprise/public customer). brought to market, Apps brought to Includes apps made available to the customer through some marketing prototyped) market (1/3) channel (e.g., in an app store, for download, or available as an enterprise solution, etc.), irrespective of revenue generated. App prototypes (1/6) Includes any functional prototype, including from hackathons, bootcamps, trainings, etc. 135 Indicator Subindicator Definition (weight) In-depth team Refers to the number of teams that received in-depth, support one-on-one support from the mLab/mHub (mentoring, coaching, etc.) Training Trained Refers to training of individuals, including over short in-depth (2/3) periods of time (e.g., one-day workshops). Includes sign-ups for training or other services for individu- Trained, any als (e.g., trainings, workshops, bootcamps, etc.). Is used as other (1/3) a proxy for individual demand for mLab/mHub services. Applications Individual Includes teams that applied for one-on-one, in-depth for support applicants (1/3) support (e.g., mentoring, coaching, etc.). Is used as a proxy for team demand for mLab/mHub Team applications ncludes teams that applied for one-on-one, in-depth (2/3) support (e.g., mentoring, coaching, etc.). Is used as a proxy for team demand for mLab/mHub services Events Includes all events (e.g., hackathons, bootcamps, seminars, workshops, Startup Weekends, Mobile Monday, etc.) that the mHub or mLab organized. Overall reach Broad measure of awareness and reach. Includes any Table C.1: developer, entrepreneur, or other client that has been in Composition contact with the mHub or mLab or shown interest in any and way (e.g., sign-up for social media or mailing lists). Definitions of Indicators Appendix D. Data Collection and Interviews The study draws from 148 semi-structured face-to-face interviews and 13 focus groups with 240 stakeholders of seven mLabs and mHubs in five countries (Armenia, Kenya, Nepal, South Africa, and Vietnam). Interviewees were asked about their perceptions on elements of the Business Model Canvas applied to the mHub or mLab in question. Interviewed stakeholder groups included develop- ers, entrepreneurs, and startups; mLab and mHub managers; partners and sponsors; policy makers; and other members of local ecosystems. Interview participants form a convenient sample; they were chosen based on the perceived relevance for the evaluation’s goals and in coordination with mLab and mHub managers. Biases in the participants’ answers cannot be ruled out, as the interviewer was clearly identified to them as an infoDev and World Bank consultant. Secondary sources included a comprehensive database of program and project documentation, including reporting material and project scorecards. The study was also informed by the findings of an independent evaluation of the CSBKE program conducted by the consulting agency Universalia.123 136 References Auerswald, P. E. and L.M. Branscomb. 2003. “Valleys of Death and ‘Darwinian Seas:’ Financing the Invention to Innovation Transition in the United States.” The Journal of Technology Transfer, 2 8 (3-4): 227-239. Autio, E. 2008. “High-and low-aspiration entrepreneurship and economic growth in low-income economies.” UNU-WIDER Workshop on Entrepreneurship in Economic Development, 21-23 August. Helsinki, Finland. Baird, R., Bowles, L., and S. Lall. 2013. “Bridging the “Pioneer Gap”: The Role of Accelerators in Launching High-Impact Enterprises.” Aspen Network of Development Entrepreneurs and Village Capital. Feldman, M., Francis, J. L., and J. Bercovitz. 2005. “Creating a Cluster While Building a Firm: Entrepreneurs and the Formation of Industrial Clusters.” Regional Studies, 39(1): 129-141. Fritsch, M. 2008. “How Does New Business Formation Affect Regional Development?: Introduction to the Special Issue.” Small Business Economics, 30(1): 1-14. H eydebreck, P., Klofsten, M., and J. Maier. 2000. “Innovation Support for New Technology-Based Firms: the Swedish Teknopol Approach.” R&D Management, 30 (1): 89-100. erner, J. 2012. The Boulevard of Broken Dreams: Why Public Efforts to Boost Entrepreneurship and L Venture Capital Have Failed—and What to Do About It. Princeton: Princeton University Press, http://press.princeton.edu/titles/8984.html. artin, R. and Sunley, P. 2003. “Deconstructing clusters: chaotic concept or policy panacea?” Journal M of Economic Geography, 3(1): 5-35. Meyer-Stamer, J. 2002. “Clustering and the Creation of an Innovation-Oriented Environment for Industrial Competitiveness: Beware of Overly Optimistic Expectations.” International High-Level Seminar on Technological Innovation. Beijing: United Nations. UNCTAD. 2012. Information Economy Report 2012: The Software Industry and Developing Countries. New York and Geneva: United Nations. http://unctad.org/en/PublicationsLibrary/ier2012_en.pdf. ong, P. K., Ho, Y. P., and E. Autio. 2005. “Entrepreneurship, Innovation and Economic Growth: W Evidence from GEM data.” Small Business Economics, 24(3): 335-350. orld Bank. 2012. Information and Communications for Development 2012: Maximizing Mobile. W Washington, DC: World Bank. http://www.worldbank.org/ict/IC4D2012. orld Economic Forum. 2013. Entrepreneurial Ecosystems Around the Globe and Company Growth W Dynamics. Report Summary for the Annual Meeting of the New Champions 2013. Geneva: 137 ©2014 infoDev / The World Bank | 1818 H Street, NW | Washington DC, 20433 Email: info@infoDev.org | Tel + 1 202 458 8831 | Twitter: @infoDev www.infodev.org