The mobile banking customer that isn’t: drivers of digital financial services inactivity in CÔTE D’IVOIRE Acknowledgment This report was written by Susie Lonie, Meritxell Martinez, Christopher Tullis and Rita Oulai, the IFC Digital Financial Services team for Cote d’Ivoire. The team would like to thank the following World Bank Group peer reviewers and colleagues for their guidance and support: Anna Koblanck, Cassandra Colbert, David Crush, Tiphaine Crenn, Corinne Riquet, Estelle Lahaye, Greta Bull, and Joseck Mudiri. We would also like to acknowledge the collaboration of two leading Côte-d’Ivoire mobile money providers in the development of this market research. The mobile banking customer that isn’t: drivers of digital financial services inactivity in Côte d’Ivoire Executive Summary 2 Background 3 Definition of DFS inactivity 3 Inactivity in context 3 Research findings: Reasons for Inactivity 5 Irregular income 6 No need to use a mobile money account 6 Too expensive 8 Access to agents 9 Service is too complicated 10 What predicts inactivity? 11 POTENTIAL WAYS TO INCREASE ACTIVITY 12 CONCLUSIONS & RECOMMENDATIONS 13 2 Executive Summary Côte d’Ivoire is the largest Digital Financial Services market 3. Too expensive: over 15 percent of customers cite cost as in the West African Economic and Monetary Union. In a reason for not using mobile money accounts instead 2014, Côte d’Ivoire accounted for over 50 percent of the of cash. Mobile money tariffs are much higher in Côte value deposited into mobile money accounts in the region. d’Ivoire than other African countries, and this certainly However, in common with many markets across the globe, seems to be restricting usage. Côte d’Ivoire has a high level of inactive DFS accounts that Some secondary reasons were also raised by up to 10 percent have not been used for over 90 days. While there are wide of the research respondents, such as a lack of conveniently differences between providers, data indicates that even located agents, difficulty using the services, and lost PIN under the best case scenario almost 50 percent of the total codes. In addition, the high incidence of “direct deposits” number of registered DFS clients are inactive. In order to (into a recipient’s mobile money account) means that sharpen the DFS business case and to further financial available data undercounts customers who use the service inclusion, it is important to understand why the services to send money. on offer are not being widely used by registered customers. Is it a matter of pricing? Could it be product design? Does it A number of recommendations can be drawn from the have to do with concerns around security? Or is it an issue of research in order to improve activity levels. Firstly, the cost customer care? of DFS transactions should be reviewed and reduced. This had strong support from inactive customers. Secondly, the To investigate the causes of inactivity and learn how they services and products of DFS providers should be made more can be addressed, The Partnership for Financial Inclusion, relevant for customers, for example by offering access to a joint initiative of IFC and The MasterCard Foundation to savings and loans. There is also a strong demand for a better advance mobile financial services in Sub-Saharan Africa, distribution of agents across the country to service more carried out market research in cooperation with two of the locations, particularly in rural areas. Finally, there is a need to largest DFS providers in Côte d’Ivoire. The study found three help customers understand the benefits of the services and main barriers to usage: how to use them with confidence. 1. Not a relevant service: nearly half of the customers interviewed have irregular incomes and therefore do not need to use their mobile money accounts on a consistent or regular basis. 2. Lack of understanding of the benefits of the service/ product: over one quarter of customers are unaware of any compelling reason to use a mobile money account instead of using cash or more traditional forms of financial services. The mobile banking customer that isn’t: drivers of digital financial services inactivity in Côte d’Ivoire 3 Background Definition of DFS inactivity and quality constraints could unlock significant additional market activity in Côte d’Ivoire. The digital financial services market is still nascent, and industry standards, definitions, and terminology have not yet been fully established. There are several definitions of Inactivity in context inactivity, but the most commonly used is “90 days inactive.” As the largest mobile money market in WAEMU, Côte It means a customer has not performed a transaction d’Ivoire accounts for just over half of the region’s registered in the past 90 days, and it is the measure of inactivity customers and number of transactions performed.4 By used throughout this report. This definition includes all the end of 2014, the country’s five digital financial services transactions where mobile money is transferred from one providers5 had together registered over 9 million mobile account to another, but does not include administrative money accounts. In the same year, Ivorians deposited over transactions such as balance enquiry or changing a PIN. FCFA 1.15 trillion, equivalent to $2.4 billion, into mobile Some other measures of inactivity exclude zero-rated mobile money accounts, which was over 70 percent of the value money transactions such as deposits and remittances deposited in the region. In terms of volume of mobile money received, but these measures are unhelpful in understanding transactions in the region, Mali and Burkina Faso follow in customer behavior as customers may be using the service second and third place (Figure 1). regularly but still not be considered active. FIGURE 1: DISTRIBUTION OF VOLUME OF TOTAL Low activity rates increase the cost of acquiring and MOBILE MONEY TRANSACTIONS WAEMU 2014 maintaining active customers, as has been argued by CGAP1: 1% 6% 0% Benin Togo Niger “If the acquisition cost per active customer 7% Senegal is $5 when a deployment has a healthy 50 percent activity rate,2 the acquisition cost is a reasonable $10 per active customer. However, if the activity rate 13% drops to 10 percent, the cost per active Burkina customer increases dramatically to $50. If 53% a deployment has an activity rate as low as Côte 1 percent – then they are paying $500 for d’Ivoire every active customer.”3 The high cost associated with high inactivity levels could in itself lead to poor agent services, as resources are diverted 20% Mali away from marketing as well as agent supervision and oversight. Good quality agents can be the most effective Source: BCEAO.6 marketing tool for DFS providers, and resolving agent supply 1 Consultative Group to Assist the Poor 2 The term “acquisition costs” refers to the cost incurred by the DFS provider in order to acquire a new customer. This terms covers 4 BCEAO (2014), “Situation des services financiers via la téléphonie commission bonuses paid to agents, any back-office know-your- mobile dans l’UEMOA.” http://www.bceao.int/IMG/pdf/rapport_ customer (KYC) requirements that must be fulfilled, the starter kit annuel_sur_les_services_financiers_via_la_telephonie_mobile.pdf for customers available with some services, etc. 5 CelPaid, Moov, MTN, Orange, and Qash Services 3 Claudia McKay (2011), “The Lurking Challenge of Activating the 6 BCEAO (2014), “Situation des services financiers via la téléphonie Inactive Customer.” http://www.cgap.org/blog/lurking-challenge- mobile dans l’UEMOA.” http://www.bceao.int/IMG/pdf/rapport_ activating-inactive-customer annuel_sur_les_services_financiers_via_la_telephonie_mobile.pdf 4 Despite these promising developments, almost 50 percent of all mobile money accounts in Côte d’Ivoire are inactive. This is not unique to the Ivorian market. According to the GSMA7, only 34 percent of the 299 million registered mobile money accounts worldwide in 2014 were 90 day active, although activity levels had increased from the previous year.8 The GSMA also estimates that across Sub-Saharan Africa, activity levels are higher than in other regions, with 61.9 million of the 146 million registered DFS customers being active, or 42 percent. Various sources of activity data as a percentage of the adult population rather than as a percentage of registered customers show big variations across Sub-Saharan Africa (Figure 2). The activity level in Côte d’Ivoire measured as a percentage of the adult population is above the average on the continent, but there is still room for improvement. FIGURE 2: ACTIVITY LEVELS OF MOBILE MONEY ACCOUNT IN OTHER MARKETS. (ACTIVE MOBILE MONEY USERS/ADULT POPULATION) 62% 29% 32% 26% 2% 2% 6% 6% 0.20% 0.20% Nigeria Benin Zambia Mozambique Cameroon Senegal Uganda Côte d’Ivoire Tanzania Kenya Source: Latest data available from various sources. Benin, Zambia, Mozambique, Cameroon, Senegal: UNCDF (2014). Nigeria: EFInA (2014). Kenya, Tanzania, Uganda: Intermedia (2013). Côte d’Ivoire: Authors’ own calculations based on World Bank activity data collected from three principal mobile providers and valid for the adult population of Côte d’Ivoire (2014). 7 GSM Association, Groupe Speciale Mobile Association 8 GSM Association (2014), “State of the Industry 2014: Mobile Financial Services for the Unbanked.” http://www.gsma.com/ mobilefordevelopment/wp-content/uploads/2015/03/SOTIR_2014. pdf The mobile banking customer that isn’t: drivers of digital financial services inactivity in Côte d’Ivoire 5 Research findings: Reasons for Inactivity The study asked a sample of one thousand mobile money customers from two leading service providers in Côte d’Ivoire why they did not regularly use their registered mobile money accounts. The three main reasons for inactivity cited by respondents were: irregular income, no need for a mobile money account, and the service being too expensive (Figure 3). Some operational causes of inactivity seen in other markets, such as agent non-performance and technical issues,9 also appear as barriers to activity in Côte d’Ivoire, but not to the same degree. FIGURE 3: MAIN REASONS FOR INACTIVITY AMONG RESPONDENTS Irregular income 43.6% No need to use it 27.0% Service is too expensive 15.5% No agents close to where I am 10.2% Service is too complicated 8.2% PIN lost 8.1% Service does not work well (technical issues) 7.2% No interoperability among wallets 4.3% Security not guaranteed 3.7% Liquidity issues 3.3% Does not know 3.2% Not offering the services I am interested in 3.2% I prefer another channel to transfer money 2.7% I do not trust the service 1.7% Customer service is bad 1.7% Other 13.8% In order to better understand the three main reasons for inactivity, the research team constructed a model to profile the respondents citing these reasons for not regularly using the service. All of the models in the following sections control for socio-demographic variables such as age, gender, and household income (see Annex 4 for details). 9 Claire Scharwatt (2013), “Should You Be Cutting Your Inactive Agents?” http://www.gsma.com/mobilefordevelopment/should- you-be-cutting-your-inactive-agents 6 Irregular income No need to use a mobile Irregular income is the main reason given for inactivity, cited money account by 43.6 percent of respondents. With irregular incomes, The second reason respondents gave for inactivity was that many customers do not have consistent use of their mobile they have no need for a mobile money account. This suggests money accounts. Customers on irregular incomes do not that DFS does not appear as a compelling alternative to cash consider themselves to have spare money to keep in their or other forms of financial services. This could partly be due mobile money accounts, suggesting that they do not see to relatively low awareness of the benefits of DFS among mobile accounts as a convenient means of serving their customers, and partly because mobile money accounts can everyday payment needs. It is unclear whether this implies still only be used for a limited range of payments in Côte that mobile money accounts are indeed unnecessary d’Ivoire. for this customer group, or if it is rather due to a lack of understanding of the benefits of digital financial services. Another reason may be the use of multiple mobile money The research survey made a distinction between unused accounts in Côte d’Ivoire. A common phenomenon in Sub- mobile money accounts that have been registered but never Saharan Africa is that customers hold multiple cellphones used, and dormant accounts that have previously been used and mobile connections to minimize cost by avoiding “off- but are currently inactive. There appears to be a number of network” calls. This phenomenon has been inflating mobile irregular users who only use their accounts occasionally, money penetration figures for many years, a problem when they have been sent money or have earned money which is exacerbated by the fact that many MNOs are now and thus have funds. Further research is needed to better automatically registering customers for their DFS when understand this customer segment and the services and a new SIM card is sold. Device sharing is also commonly products that are relevant to them. observed, having a similarly distorting effect but leading to an undercount of mobile phone usage instead. Close to 95 The profile of those responding that irregular income is a percent of respondents in this survey reported two or more reason for inactivity is as follows: cellphones per household, and most respondents had at least two SIM cards (Figure 4). • Receiving money. People who use their mobile money accounts to receive funds are approximately 25 percent On average, respondents had 1.6 mobile money accounts, more likely to report lack of funds as a reason for indicating that they may use different digital accounts for inactivity. different reasons, specific use cases, or due to poor network • Being a student. Students are twice as likely to report lack coverage. Another contributor to multiple mobile money of funds as a reason for inactivity. accounts may be the entry of a third market actor in 2013, • Education. People who have at least some secondary an MNO that heavily promoted free self-registration of education are 40 percent less likely to report lack of mobile money accounts via a USSD code, which could have funds as a reason for inactivity. increased the registration of multiple accounts. It seems • Household income. People in poorer households are more likely that customers register for several mobile money likely to report lack of funds as a reason for inactivity.10 accounts, but only end up using the preferred service. 10 No other variables in the model (including demographic controls such as age, gender, and whether respondents live in rural or urban communities) significantly explained why respondents reported irregular income as a reason for e-wallet inactivity The mobile banking customer that isn’t: drivers of digital financial services inactivity in Côte d’Ivoire 7 FIGURE 4: PERCENTAGE OF RESPONDENTS WITH MULTIPLE SIMs 49.7% 26.3% 19.0% 4.6% 0.4% 1 2 3 4 5 The profile of those indicating that they are inactive because account to send money to other individuals (Figure 5). This they do not have a need for a digital account is similar to that suggests that those customers that use their accounts to of customers who stated irregular income as a reason for save and receive money (as opposed to sending money) may, inactivity: in some cases, only be occasional users who are not taking full advantage of the range of services that mobile money • Usage. Both those who receive money and those who use provides. It also suggests that the majority of transfers their mobile money account to save are approximately may be emanating from a relatively small number of highly 50 percent more likely to report lack of need as a reason active benefactors. This situation would leave relatively for inactivity. few opportunities for the accounts of users who are • Being employed. Being employed makes one half as likely primarily receivers and savers to be funded, explaining why to report a lack of need as a reason for inactivity. those accounts would go inactive. Encouraging increased • Education. People who have at least some secondary numbers of senders would thus probably increase the education are 20 percent more likely to report lack of activity numbers of those who primarily receive and save need as a reason for inactivity. money, which would also allow them to take advantage of • Household income. People in poorer households are more other digital financial services. likely to report lack of need as a reason for inactivity.11 To better understand why people would open an account One important factor influencing the data on customers’ that they would then report having no need to use, it is usage patterns is the phenomenon of direct deposits. A instructive to review the data on how respondents use their direct deposit occurs when the customer hands cash to an mobile money accounts. The above model shows that two agent and initiates a deposit, but gives the agent the mobile usage behaviors – saving and receiving money – significantly number of the recipient instead of his or her own number. predict whether or not customers will report lack of need The agent thus deposits the cash directly into the recipient’s for their mobile money account as a reason for inactivity. account, allowing the sender to circumvent the electronic These were also the two usage behaviors reported by the transfer process and avoid paying a transfer fee. This direct highest proportion of users (with 63 percent reporting using deposit behavior has become the norm in Côte d’Ivoire rather their accounts to receive money and 47 percent reporting than the exception, as is the case in many other markets. In using them to save). The remaining usages were reported our sample, 31 percent of users reported having used the by a significantly smaller percentage of users. For example, direct deposit method to transfer money in the past, while only 11 percent of users reported using their mobile money only 11 percent of users reported having made an electronic transfer using their mobile money account. 11 No other variables in the model (including demographic controls such as age, gender, and whether respondents live in rural or urban communities) significantly explained why respondents reported a lack of a need to use their e-wallet as a reason for inactivity. 8 FIGURE 5: HOW WAS THE MOBILE MONEY ACCOUNT USED IN THE PAST? 63% 47% 31% 20% 15% 11% 6% 5% 4% 1% 0% Receive Save Deposit to None Buy Deposit Pay school Pay bills Other Pay Does not money 3rd party airtime to 3rd fees merchants know via agent party via transfer Direct deposits are problematic for a number of reasons. For revenue may reduce providers’ incentive to invest further one, they give a false reading of inactivity. Since a sender does in the service. It also provides a disincentive for users to not use his or her own mobile money account to transfer deposit money into their own accounts and store it there money to the recipient, the sender in such a transaction is to be transferred later, since it is cheaper to store money as technically not active, despite enjoying the benefits of the cash and go to an agent in person when they need to make service. Indeed, a sender does not need to be a registered a transfer. customer or even own a mobile phone to transfer money through direct deposit. For MNOs that are DFS providers, Too expensive this means that the strategic benefits that digital accounts may bring to their core business in terms of customer High cost is a key constraint to usage for 15.5 percent of acquisition and retention are less significant. respondents. In the case of Côte d’Ivoire specifically, the respondents may have a point. CGAP compared the cost for Another negative consequence of direct deposits is that the a customer to use a bundle of digital financial services from DFS provider has to pay the agent for the deposit, which is 15 providers in selected African and Asian markets and found free to the customer, but does not receive any income from that the two Ivorian DFS providers tied in the position of the person-to-person transaction that would have occurred fourth most expensive of that sample (Figure 6). High prices if the money had been sent from one mobile money account also drive the undesired “direct deposit” behavior. to another as intended. The resulting loss of expected The mobile banking customer that isn’t: drivers of digital financial services inactivity in Côte d’Ivoire 9 FIGURE 6: AVERAGE PRICING TO USE A BUNDLE OF MOBILE MONEY SERVICES IS USD 3.9 PER MONTH 8.2 6.3 4.3 4.4 4.4 4.4 4.9 3.7 4.1 2.3 2.8 2.9 2.2 1 1.5 Zap (KN) EKO (IN) Gcash MPESA WING Smart Caixa Vodafone Bradesco MTN M-Paisa Orange Zap (TZ) WIZZIT easypaisa (PH) (KN) Money Money (BN) M-PESA (BN) Mobile (AF) money (ZA) (PK) (CAM) (PH) (TZ) Money (CI) (CI) Source: CGAP. 12 Note: All prices are given in international dollars (adjusted for purchasing power parity) in order to facilitate comparisons across markets. The following characteristics significantly predicted which Access to agents users reported price as a reason for inactivity in our sample:12 The fourth most cited reason for inactivity is a lack of agents • Usage. People who used their mobile money account located close to the customer. The agent is the customer’s to transfer money to individuals or to pay school fees key contact point and has a critical influence on usage of are 70 percent more likely to report price as a reason for the service. Common issues at the agent contact point that inactivity. The opposite trend is true for those who used could influence activity levels include lack of liquidity (either the digital accounts to buy airtime. Such customers are mobile money or cash), lack of knowledge about how to use 50 percent less likely to report high fees as a reason for the service or advise customers on the use of the service, and being inactive. a physical lack of agents where they are needed, especially • Education. People who have had at least some secondary in rural areas. In this survey, customers in rural areas were education are 67 percent more likely to report expense 3.25 times more likely to report lack of agents in the area as a as a reason for inactivity. This suggests that people with reason for inactivity, controlling for other variables. more education may be more price sensitive. • Household income. People in poorer households are more A balance is required between the proximity and the likely to report price as a reason for inactivity.13 productivity of agents. There needs to be enough agents in Respondents for whom price was a key reason for inactivity an area to service all customers, but there is also a risk of tended to be senders rather than recipients of mobile money, overcrowding agents, making the individual agent business which makes sense given that storing and receiving money case unviable. In Kenya, for example, the ideal number of in a mobile money account is free while there is a fee for active customers per active agent was found to be 600.14 making a transfer. In 2014, Côte d’Ivoire had the highest number of registered agents, 18,000, and the highest ratio of registered DFS customers per agent in the WAEMU region, 483 versus a 12 Claudia McKay and Mark Pickens (2010), “Branchless Banking 2010: Who’s Served? At What Price? What’s Next?” http://www.cgap.org/ sites/default/files/CGAP-Focus-Note-Branchless-Banking-2010- Who-Is-Served-At-What-Price-What-Is-Next-Sep-2010.pdf 14 Neil Davidson and Paul Leishman (2012), “Building, Incentivizing 13 No other variables in the model (including age, gender, and and Managing a Network of Mobile Money Agents: A Handbook employment status) significantly explained why respondents for Mobile Network Operators.” http://www.gsma.com/ reported price as a reason for inactivity. mobilefordevelopment/wp-content/uploads/2012/03/building.pdf 10 regional average of 220.15 It would appear that the supply of respondents rated the ease of use of mobile money accounts agents is currently sufficient as a lower ratio of customers and their actual ability to use these accounts unaided. to agent may reduce agent productivity, assuming that Relatively few customers agreed when asked if they found all agents are currently efficiently run. However, as lack of it difficult to use their digital account, but when asked access to agents has been cited as one of the key causes whether or not they need help from others to use the mobile of inactivity, it would appear that some current agents are money service, 35 percent said they do (Figures 7 and 8). This not situated where they are needed, are not active, or both. gap suggests that customers may be reluctant to admit that More research on the location of agents and their level of they have difficulty using the service, perhaps due to social activity is needed to determine whether the existing mobile stigma attached to being illiterate, and that customers thus money agent network is sufficient to meet existing and underemphasize the extent to which interface complexity is future demand. a barrier to activity. FIGURE 7: IS THE MOBILE MONEY SERVICE EASY TO To better understand some of the aspects of agent network UNDERSTAND/USE? management which can be improved to increase activity 17.7 % levels, the research survey was complemented by 40 mystery shopping visits to existing agents. Six key dimensions of the customer experience at the agent were assessed: 1. External environment at agent location 2. Internal environment at agent location 3. Waiting and dispatch time 4. Customer service 5. Agent’s knowledge of products 6. Overall agent behavior From the mystery shopping scorecards it is evident that one area that needs further improvement is the internal 82.3 % environment at agent location. Issues like security within Yes No the agent location and lack of flyers and communication materials were mentioned as very problematic. In general, FIGURE 8: DO YOU NEED HELP TO USE MOBILE MONEY SERVICES? the mystery shopping exercise indicated that the guidance given or help offered by staff at agent locations is very poor for the sample analyzed. 35.21 % Service is too complicated Mobile money accounts seem easy to use, but a third of respondents said they need help to use the service. Only 5 percent cited this as a reason for inactivity, however. This apparent paradox is due to a discrepancy between how 64.79 % 15 The BCEAO does not (yet) track whether agents are active or not. BCEAO (2014), “Situation des services financiers via la téléphonie mobile dans l’UEMOA.” http://www.bceao.int/IMG/pdf/rapport_ annuel_sur_les_services_financiers_via_la_telephonie_mobile.pdf No Yes The mobile banking customer that isn’t: drivers of digital financial services inactivity in Côte d’Ivoire 11 This suggests that there are hurdles to the use of mobile sometimes appear to use their mobile money accounts as money accounts that customers are either not aware of or bank accounts. do not want to admit to. It is also possible that registered customers understand in theory what they need to do, but The finding that customers who use their mobile money that in practice the user experience is not as easy as Figure 8 accounts to buy airtime are less likely to be inactive is key for seems to indicate. Practical barriers may include issues such DFS providers. Currently, relatively few customers use digital as ensuring that the correct destination number is entered accounts to buy airtime, with 8 percent of respondents or USSD time-out. reporting doing so regularly, while 62 percent had never used a mobile money account to buy airtime (Figure 9). Although the number of airtime purchase has been increasing over What predicts inactivity? the last years, there is still significant untapped potential A key objective of this research study is to help DFS providers to boost mobile money activity through marketing efforts understand how to activate customers. For that reason, the focused on airtime purchases. team created a simple model to identify and examine the socio-economic characteristics of inactive customers (see annex 4 for details). Four of the variables included in the FIGURE 9: PURCHASE OF AIRTIME VIA MOBILE ACCOUNT model were found to be significant predictors of inactivity.16 7.6% 1. Being employed. Customers that are employed are 75 percent less likely to be inactive than students and the unemployed. 14.4% 2. Usage. Those buying airtime via a mobile money account are approximately 50 percent less likely to be inactive, and people depositing money directly into a third party’s account at an agent are 35 percent less likely to be inactive. 15.8% 3. Being unbanked. People who have no traditional bank 62.2% account are 50 percent less likely to be inactive than those that are banked. 4. Low service satisfaction. Customers that are unhappy with the mobile money service are highly likely to be inactive. Never Rarely Sometimes Often The fact that clients who are not otherwise banked were less likely to be inactive than banked clients suggests that some people may use mobile money accounts as substitutes for bank accounts. Indeed, 47 percent of respondents had used their digital account for saving at some point, and 70 percent planned to do so in the future. Further research is needed to better understand if, how, and why customers 16 No other variables in the model (including demographic variables such as age and gender) significantly explained inactivity. 12 POTENTIAL WAYS TO INCREASE ACTIVITY This study included focus groups that explored a number of initiatives or changes that may help to increase activity levels among customers. Focus group participants were asked what DFS providers could do to make them use their mobile money account. Not surprisingly, lower prices scored as the main reason. The second solution suggested was to extend the range of services, indicating that the current offering may not be fully catering to the needs of the Ivorian market (Figure 10). FIGURE 10: SOLUTIONS SUGGESTED BY RESPONDENTS TO INCREASE ACTIVITY RATES 20.9% 20.6% 15.2% 13.5% 10.5% 8.3% 7.0% 3.9% Decrease prices Extend number Improve the Simplify Improve Increase Reassurance of Other of services distribution registration communication e-commerce via the security of network mobile money service As highlighted in a recent publication by IFC and The Customers were also asked for their opinion of a range of MasterCard Foundation, users of digital financial services in potential initiatives that could increase usage of digital Côte d’Ivoire desire a storage mechanism that allows them financial services (see appendix 3 for details). The three to keep their money safe. It appears that many do not 17 preferred initiatives were related to rewards, access, and believe that mobile money accounts offer them this service, products: or more likely, they have not thought about it as this kind of 1. The DFS provider offers bonuses to clients. service as it is not marketed as such. Therefore, increasing 2. The client can access mobile money accounts via ATMs mobile money’s viability as a savings solution could increase and bank-to-mobile account solutions. its relevance for customers and decrease inactivity. The third 3. The DFS provider offers microcredit via mobile money solution to inactivity suggested by customers is to improve accounts. the distribution network. A thorough diagnostic of each provider network is needed to disentangle the reasons why Some of these suggested solutions are already present in the clients appear unhappy with the existing distribution of Ivorian market: DFS providers often offer airtime bonuses digital financial services. for bill payments or airtime purchases via mobile money accounts. Some mobile money operators also offer flat fees for local and international person-to-person payments on a 17 Susie Lonie, Meritxell Martinez, and Rita Oulai (2013), “Overview of Côte d’Ivoire Mobile Financial Services Market Data 2013.” http:// www.ifc.org/wps/wcm/connect/60efa900461a20518634bf991618 2e35/ifc_e-money_english.pdf?MOD=AJPERES The mobile banking customer that isn’t: drivers of digital financial services inactivity in Côte d’Ivoire 13 CONCLUSIONS & RECOMMENDATIONS temporary basis to encourage repeated usage. Two of the There is no single solution to increasing customer activity DFS providers have been experimenting with ATMs: one by in the digital financial services market. However, there are leveraging the large existing ATM network of a state-owned a range of actions that can be taken to help overcome the bank, and the other by investing in and deploying its own issues that contribute to an inactive clientele. This study ATM network. The integration of bank accounts with mobile has highlighted some important new areas for research and money accounts already exists in Côte d’Ivoire, allowing lessons learned for the industry as a whole: for the transfer of money from digital accounts to bank 1. Products and services could be developed that better cater to accounts and vice versa. However, this service is fairly recent, customers with irregular incomes. Irregular income is the quite expensive, marketed only on a limited scale and targets main reason for inactivity cited by registered customers. existing bank customers. This service, if competitively priced The Ivorian economy is strongly dependent on the and promoted, could also be used by agents to manage their primary sector and the informal sector, with 68 percent liquidity. of the workforce employed in agricultural activities with seasonal incomes.18 Mobile money account activity, The demand for more DFS-enabled microcredit products however, is measured over a period of 30 days, or, highlighted in focus groups is not yet being fully explored more commonly, 90 days. Research is needed to better by Ivorian providers. Product innovation is led by markets in understand the reasons, prevalence, and impact of East Africa, with M-Shwari in Kenya, M-Pawa in Tanzania, irregular usage of digital financial services, and possible and Igurize Amafaranga, an Airtel microcredit product measures to take account of this in order to serve recently launched in Rwanda. There is an opportunity to customers with irregular or cyclical cash flow, such as provide similar services in Côte d’Ivoire, where recent data farmers and other seasonal workers. from BCEAO confirm that the DFS ecosystem in the WAEMU 2. DFS prices should be revised and reduced. The pricing of region is already being used for safe, non-remunerated mobile money services appears to be a major issue in storage of funds. In partnership, financial institutions and Côte d’Ivoire. It is unclear why the Ivorian providers DFS providers could leverage mobile money’s high market charge higher rates than in many other markets, but this is certainly limiting the growth potential of their services. penetration to bring interest-bearing savings products and Lower prices might encourage more active usage, and credit services to the unbanked in Côte d’Ivoire. promotional or pilot initiatives by the providers are recommended to determine the cost versus benefit of price reduction over an extended period. 3. Customer education could be improved. Lack of customer information and education is an issue both at agent locations and as demonstrated by how many clients need help using the service. There is a clear need to give customers a better understanding of the uses of mobile money accounts, and to disentangle what the barriers to ease of use are and how these are linked to behavioral, contextual, or even personality-related factors. 18 IFC (2013), “Scan du secteur agro-business Côte d’Ivoire.” 14 4. A solution to direct deposits is needed. Direct deposits are 6. A wider range of products and services could be explored. The very common in Côte d’Ivoire, negatively impacting the range of uses to which mobile money accounts can be DFS provider business case and creating a false measure put in Côte d’Ivoire is growing, but there is still significant of inactivity. Eliminating revenue leakages due to direct room to develop more and better services for customers deposits could also provide a means of financing fee (Figure 11). The market is currently at the stage of decreases. becoming established and is ready to move into offering 5. Agent availability and management can be improved. value added services and third party developments, such Agent access and performance have a strong impact as interest-bearing savings, loans, merchant payments, on customers’ usage of digital financial services. Good microfinance integration, and micro-insurance. There quality agents are a major asset, whilst poor agents can is also significant scope to increase usage of existing put off customers from using the service. In Côte d’Ivoire, services, for example by encouraging more billers to the number of agents is reported to BCEAO but it is not accept DFS payments or by promoting the use of airtime known what proportion is active. 19 GSMA estimates purchase via mobile money accounts. In addition to that, on average, 60 percent of agents were active increasing usage and activity, network operators that worldwide, and 43 percent in West Africa, at the end are DFS providers could benefit from decreased costs if of 2014.20 Further research on agent networks in Côte they can increase sales of airtime through mobile money d’Ivoire is needed to understand where the gaps and accounts. issues lie, both in terms of proximity and productivity. FIGURE 11: MFS SERVICE EVOLUTION MFS SERVICE EVOLUTION THIRD PARTY CUSTOMIZED PRODUCTS ON SALE – INSURANCE; LOYALTY; HIRE-PURCHASE; CREDIT RATING; ETC DEVELOPMENTS LINKED POS CARD MERCHANT PAYMENTS LOANS INTEGRATED SERVICES INTEREST EARNING SAVINGS SALARY / B2C/ G2P Côte d’Ivoire BILL PAYMENT SAFE STORAGE AIRTIME PURCHASE BASIC SERVICES P2P TRANSFERS CASH IN / CASH OUT NEW / DEVELOPING ESTABLISHED MATURE INCREASING SERVICE SOPHISTICATION Source: Susie Lonie 19 The Alliance for Financial Inclusion suggests the following definition of “active MFS agent”: “MFS cash points that have performed at least one transaction in the last 90 days.” AFI Mobile Financial Services Working Group (2013), “Mobile Financial Services Indicators for Measuring Access and Usage.” http://www.afi-global.org/sites/ default/files/publications/mfswg_guideline_note_no.11_lo.pdf 20 GSM Association (2014), “State of the Industry 2014: Mobile Financial Services for the Unbanked.” http://www.gsma.com/ mobilefordevelopment/wp-content/uploads/2015/03/SOTIR_2014. pdf The mobile banking customer that isn’t: drivers of digital financial services inactivity in Côte d’Ivoire 15 Annex 1: Annex 2: Research approach Mystery Shopping Methodology The research involved both quantitative and qualitative data Mystery shopping with from the two DFS providers, using the following structure: structured questionnaire (PAPI METHODOLOGY method) and at least two visits per distribution point. • Step 1: Data analytics on over 700,000 randomly chosen DFS customer accounts21 • Step 2: Face-to-face interviews with 50 inactive DFS Branches and point of sales TARGET customers managers, agents of the two POPULATION participating MNOs • Step 3: Phone-based survey with a random sample of 1,000 DFS customers22 • Step 4: Mystery shopping trips to 40 DFS agent outlets • Step 5: Eight focus groups with inactive DFS customers 40 distribution points TOTAL SAMPLE representing all distributor types All field work in step 3 was split between Abidjan, San Pedro for each MNO. and Bouake regions and where possible, urban, peri-urban and rural samples were used. SAMPLE DEMOGRAPHICS % DATES October 2 to 23, 2014. Men 74 Women 26 Rural 15 Urban 85 Number of visits per locality Banked 44 Unbanked 56 LOCATION* NUMBER OF VISITS Income under CFA 300,000 per month 76 Abidjan 44 Income over CFA 300,000 (or no answer) 23 San Pédro 18 Under 30 years old 44 30-45 years old 45 Bouaké 20 Over 45 years old 12 Total 82 ACTIVITY LEVEL OF THE 1000 SAMPLE *Mystery shopping conducted in main city and surrounding 28% 25% areas representing a mix of urban, semi-urban and rural 23% 20% areas. 4% Active in the last 30 days Active in the last 31-90 days Inactive for more than 90 days Have never used their ewallet Don’t know 21 The data analytics drew fewer results than expected. In 2012 the Government of Côte d’Ivoire enacted a data privacy law that prevented MNOs from sharing socio-economic information of individuals with IFC. While both MNOs shared two datasets of inactive customers, the data provided related to their MSISDN, date of subscription, type of account and revenues for a fixed period. These were insufficient data points to disentangle the socio- demographic profiles of inactive customers. 22 Participant MNO1 shared a total subsample of 100,000 customers. MNO2 shared a database with 611,740 customer files. Out of those files, 10,000 customers were selected randomly from each MNO. The research team drew a subsample of 1000 customers for phone interviews (5031 calls were conducted to obtain the 1000 sample). 16 Annex 3: Concepts Tested during Focus Groups CONCEPT TESTED IN DEFINITION FOCUS GROUPS Bonuses The MNO offers you two new benefits:   Option 1 - Any subscriber who makes more than two transactions (excluding deposit and balance inquiry) per month gets 10% airtime bonus on cash out. Option 2 - For three withdrawals charged to the receiver over one month, the third withdrawal is offered. Option 3 - For two bill payments made ​​on two successive months, the third bill payment is offered. Mobile Agents For your mobile money operations, the MNO guarantees the presence of an agent at least one fixed day of the   week (chosen by participants) in your village or in a point close to your home. Mobile Money Voice The MNO wants to launch a voice service that allows you to use the mobile money services or to make Support/IVR transactions following the instructions of a voice server. You only have to follow the IVR instructions for your   transactions. For example, you want to transfer money and you are not familiar with the process. This voice service supports you and you only have to enter the recipient’s number and amount to be sent. Young Ambassadors The MNO wants to promote its first young ambassadors by allowing them to sponsor relatives or friends. For any new subscriber enrolled and who performed a first deposit, the MNO offers a percentage (to be defined by participants) of this first deposit.   The MNO offers its subscribers two new convenient ways to deposit and withdraw from their wallets: ATM and Bank-to- • The first is to withdraw from ATMs located in its branches. The customer then perform the transactions wallet from a screen by entering his/ her PIN code. The customer can withdraw money from the ATM in the same way as at banks’ ATMs. • The second option is offered to banked customers allowing them to transfer money from their bank account to their mobile wallet. Microcredit The irregularity of cash and revenue can be an obstacle to active mobile money usage. The MNO offers to provide convenient microcredits of up to 15,000 CFA ($30) without going to an agent or filling a form. This microcredit is reimbursable over a month at a rate to be determined by the participants. Annex 4: Description of the statistical models used The analysis of the reasons for inactivity was conducted using maximum likelihood models (binary logistic regression) using four different dependent variables. The first three dependent variables capture whether or not clients reported the following factors as reasons for inactivity: 1) irregular income, 2) lack of need to use the service, and 3) the service being too expensive. The fourth model used self-reported inactivity as the dependent variable, with users being considered inactive if they reported using their mobile money account at an average frequency of once every three months or less. All four models employed similar explanatory variables, including a) how clients use their mobile money accounts (saving, paying bills, etc.), b) clients’ ability to manipulate the mobile money account interface without help, c) whether or not clients have a formal bank account in addition to their mobile money account, and d) overall client satisfaction with their mobile money accounts. The models also control for various demographic variables, including client education level, employment status, household income, gender, age, and whether clients live in rural or urban communities. Authors SUSIE LONIE is a Digital Finance Specialist working with IFC. Susie is one of the creators of the M-PESA money transfer service. In 2010 Susie was the co-winner of “The Economist Innovation Award for Social and Economic Innovation” for her work on M-PESA. MERITXELL MARTINEZ is an IFC Operations Officer and has been with IFC for 3 years. She leads a program that seeks to increase digital financial services’ uptake in Côte d’Ivoire . Meritxell is based in Dakar and works on microfinance and digital finance advisory services and portfolio management in West Africa. RITA OULAI is an IFC Operations Analyst consultant based in Abidjan. She works in microfinance and digital finance. Rita worked previously for GIZ in Côte d’Ivoire and holds a Business Administration degree from the State University of New York at Buffalo. CHRISTOPHER TULLIS is a research consultant with both the IFC Digital Financial Services team and the World Bank’s Africa Region Gender Innovation Lab. Christopher is based in Abidjan and works on the barriers to and the effects of expanding access to savings products for the poor. The PARTNERSHIP FOR FINANCIAL INCLUSION is a joint initiative of IFC and The MasterCard Foundation to expand microfinance and advance digital financial services in Sub-Saharan Africa. The Partnership is also supported by the Bill & Melinda Gates Foundation and the Development Bank of Austria (OeEB, Oesterreichische Entwicklungsbank AG), and collaborates with knowledge partners such as the World Bank and the Consultative Group to Assist the Poor (CGAP). An important objective of the Partnership is to build and share industry knowledge for the public good. This publication is part of a series of research reports published by the program. www.ifc.org/financialinclusionafrica CONTACT Anna Koblanck Communications Officer 14 Fricker Road, Illovo 2196 Johannesburg, South Africa akoblanck@ifc.org