94909 PARTNERSHIPS IN MOBILE FINANCIAL SERVICES 1 IFC Advisory Services in Sub-Saharan Africa Partnerships in Mobile Financial Services: Factors for Success Mark Flaming, Aiaze Mitha, Michel Hanouch, Peter Zetterli, Greta Bull ACKNOWLEDGMENTS This paper is the product of close collaboration between a number of individuals and institutions. The authors would like to thank Janine Firpo for her contributions to the development of the initial concept and Brad Jones for his work on the WING case study. CGAP provided manpower, financial and intellectual support, and was a valued knowledge partner throughout the process. The MasterCard Foundation has been an important supporter of both IFC’s and CGAP’s knowledge and learning work, and we are grateful for their on-going support and advice. Finally, the team would like to acknowledge Equity Bank, MTN Ghana, Tameer Microfinance Bank, Telenor Pakistan and WING Cambodia for their participation and meaningful contributions to this research. Without their willingness to share their experiences, the paper would not have been possible. Contents EXECUTIVE SUMMARY 4-5 INTRODUCTION 6 1. THE STRUCTURE OF MFS IMPLEMENTATIONS: FOUR KEY ROLES 7 2. THE CASE STUDY IMPLEMENTATIONS Equity Bank Kenya: when banks and MNOs compete 8 WING Cambodia: when the core business evolves 9 Easypaisa in Pakistan: the long-term view 10 MTN Ghana: when the leader is forced to follow 11 3. ALIGNING COMPETITIVE FORCES, ECONOMIC MOTIVATION AND ROLES 12-13 4. THE CORE REVENUE SOURCES IN AN MFS SUPPLY CHAIN AND THE DISTRIBUTION OF REVENUE 4.1 Payment Services 14 4.2 Banking 15 4.3 Telecommunications 16 4.4 The Agent Network 16 4.5 Revenue Distribution 17 5. FINAL OBSERVATIONS 18 AUTHORS 19 ACRONYMS ANM Agent Network Manager ARPU Average Revenue Per User ATM Automated Teller Machine B2B Business-to-Business MFS Mobile Financial Services MNO Mobile Network Operator NPL Non-Performing Loan OTC Over The Counter P2P Person-to-Person POS Point-of-Sale PSP Payment Service Provider SMS Short Message Service STK SIM Toolkit USSD Unstructured Supplementary Services Data 4 PARTNERSHIPS IN MOBILE FINANCIAL SERVICES Executive summary Mobile financial services have attracted significant attention in recent years as a potential means of accelerating access to financial services for the poor. Safaricom’s M-Pesa operation in Kenya demonstrated that payments services - and more recently banking services through the M-Shwari partnership with Commercial Bank of Africa - can be extended to the mass market in a cost-effective and profitable manner. Many new implementations have sprung up since M-Pesa’s launch understanding what makes partnerships work. The first is that there and regulators have responded by trying to shape the landscape for are a variety of competitive forces that create the impetus for any financial inclusion in their markets. As a result of regulatory changes MFS implementation. Understanding the competitive forces that in particular, most mobile financial services (MFS) implementations create the incentives to launch an MFS implementation is a starting today can be characterized as partnerships. Regardless of whether point for understanding the commercial motivations behind the partnerships are driven by economic or regulatory necessity, they initiative, the core business model, the players, and any partnerships have become an important feature of MFS delivery in many they might create to achieve their business objectives. The second markets. The purpose of this paper is to explore the experience concept is that there are at least four core businesses that can and with partnership models in four cases, and to extract lessons learned typically must generate revenue in the supply chain: the payment for the broader MFS industry. service business, banking, telecommunications and the agent Partnerships are an important element of MFS implementations, network. The third concept is that partnership agreements must but can vary considerably in their intent and construction. distribute revenues to ensure that the implementation sustains all Partnerships succeed or fail depending on alignment around three companies in the supply chain. This study examines four MFS high-level concepts, which together provide a useful framework for partnerships using this framework. PARTNERSHIPS IN MOBILE FINANCIAL SERVICES 5 Key findings • An array of factors appears to undermine the effectiveness of rather than a longer-term view that also encompasses benefits MFS partnerships. The deficiencies typically manifest in some generated for the partners’ core businesses, which may require combination of two forms: one or more of the partners is a more patient strategy. not playing a role that is key to their success, and/or one or • Partnership roles in an MFS implementation must be aligned more of the partners is playing a role they are ill-equipped or with competitive advantage and motivation. The motivation unmotivated to play. In some cases, the presence of what appear and ability of companies to play their roles are largely a to be suboptimal arrangements may simply be related to the function of the competitive dynamics in the banking, mobile early stage of the industry and the inevitable learning curve of communication, payment services and agent-based distribution first movers. The rapid evolution of the MFS industry is likely industries. In any given market, companies in these industries another force. The case study implementations demonstrate will have relative strengths and economic motivation to that markets, technology and business models can evolve operate different parts of an MFS supply chain. Partnerships rapidly and significantly in this sector in relatively short periods can add value where competencies and business models are of time. Partnership arrangements that appear ideal at the start complementary; but partnerships may not be possible where may become problematic as the business changes. companies have competing interests to control some part of • To be successful, MFS partnerships must enable the partners the supply chain or some service component. The challenge to generate value for their respective companies. Some part can begin with the fundamental question of which customers of that value will come from the MFS business itself, and this to acquire. Banks, MNOs, PSPs and ANMs may have different will be a function of how efficiently the partners can play views on which MFS customers are most likely to benefit the their respective roles to generate value for customers with the respective core businesses. MFS channel. Additional value will accrue to the partners’ • Finally, regulatory restrictions are the most consistent cause core businesses. For example, banks may benefit from deposit for suboptimal partnership arrangements in the four case mobilization, cost savings, and product line growth; mobile studies. In some markets, non-bank players will have more network operators (MNOs) may benefit from an increase in motivation and competitive advantage than banks, and bank- airtime sales, reduced distribution costs and client retention; based regulation can deny these companies from leading an and payment service providers (PSPs) and agent network MFS implementation. In other markets, banks with a mass managers (ANMs) may derive benefits from new customers retail ambition may have strong motivation to lead an MFS in their networks and growth in transactional volumes. The implementation, but the absence of appropriate agent banking division of revenue and cost between partners, as well as the regulations may place them at a disadvantage. To support timing of accrual of value to the core business of each partner, financial inclusion, leveling the playing field for different types plays an important role in the evolution of the implementation. of institutions in terms of key factors like issuing e-money, Many implementations have taken a short-term view that identifying and using agents and accessing communication focuses on the value generated by the implementation itself, channels should be a priority for regulators. 6 PARTNERSHIPS IN MOBILE FINANCIAL SERVICES Introduction Mobile financial services implementations are structurally complex, typically requiring expertise in banking, telecommunications, technology, marketing and distribution. Rarely will one company have the core competence to perform all of these functions efficiently, although it is not uncommon to see one player fulfilling multiple roles. In addition, banks are often required by regulation to play certain roles, and mobile network operators must provide communications services. As a result, MFS implementations typically involve some in Kenya, MTN in Ghana, WING in Cambodia, and Easypaisa in combination of financial institutions, mobile network operators, Pakistan. We explored various aspects of these implementations to agent networks managers and payment service providers.1 These extract lessons about how partnerships position the participating different companies enter into business relationships to link all of companies to jointly deliver an attractive customer value proposition the components into a seamless service delivery channel. while at the same time generating value, and distributing that value among themselves. While the findings are not easily distilled into This paper provides an examination of partnerships, a critical a checklist for successful partnerships, this examination of four aspect of MFS implementations that is particularly difficult to case studies provides some useful reference points for building get right. Companies that are very successful in their respective productive partnerships in MFS implementations. banking, communications, and payment services businesses agree on suboptimal partnership arrangements often enough to merit an This paper is organized as follows: investigation of how and why this happens. This paper explores these relationships on the assumption that getting them right is • Section 1 briefly presents the structure of MFS implementations key to the success of any MFS implementation. More specifically, and in particular the four key roles which must be fulfilled. this paper focuses on business relationships in which companies • Section 2 describes the case study implementations and the partner to create the core business and deliver the customer value partnership arrangements between the key actors. proposition in an MFS implementation. This implies a distinction from a mere contractual relationship with a service provider: • Section 3 describes the importance of aligning competitive partners share risk or depend on mutual performance of defined forces, economic motivation and partner roles. roles for the success of the implementation. • Section 4 presents key lessons related to the four underlying The research for this paper is based largely on a review of businesses of an MFS implementation, distinguishing between the four emerging MFS implementations led by different types of direct and indirect revenue that can accrue to the core businesses institutions, each with unique partnership arrangements with of the partner organizations. It also presents key lessons related other companies that operate parts of the channel: Equity Bank to the distribution of this revenue between partners. 1 In this paper, cash agent networks are a core feature of mobile financial service implementations, which distinguishes them from mobile channels that simply provide mobile phone based access to bank accounts in an additive rather than transformational manner. PARTNERSHIPS IN MOBILE FINANCIAL SERVICES 7 1. The structure of MFS implementations: four key roles MFS implementations are structurally complex. They typically require safe storage of funds, a reliable and widely available communication channel, effective marketing and delivery of payments, and the ability to put money in and take money out of the system through a widely distributed agent network. Partnerships are what hold the core pieces together. To understand this function, it is useful to characterize the • Delivering the payments service economic structure of MFS implementations, and in particular This is the company that brands and sells the payments service the four key roles that providers must partner to deliver. to the public. It may be a bank, a mobile network operator, or Companies have many options for how to distribute the different a third-party payment service provider. roles in an MFS implementation, how to contract with each other • Safely storing funds to coordinate these roles, and how to share revenue. Companies may play any combination of roles, but at least four core functions The individual accounts may be in a bank or an e-wallet are typically present in any implementation (see list on the right). platform, but ultimately the funds are stored in a bank.2 A bank or banks may also provide banking services through the Functions in the supply chain should not be confused with partners channel. Banks will also typically settle any transactions that in the implementation. Individual partners can fulfill multiple occur with accounts outside of the service network. functions: for example, an MNO can act as the payment services provider, the agent network manager and the communications • Providing a secure, cost-effective and widely channel. A bank can be the PSP, the ANM and the float holder. Or available communication channel a third party can be the PSP and ANM, leaving the bank and the Real time transactions conducted at agents or on customer MNO to fulfill only their core functions in communications and phones will pass across communication channels (e.g. USSD, banking. In all cases, a bank must play the role of holding the float SMS or data) provided by MNOs. And MNO airtime may be account and the MNO must provide the communications channel. sold across the channel. Other than this, various permutations are possible. The important point is that each of these roles needs to be filled to ensure the • Facilitating getting cash-in and cash-out of functioning of the supply chain. the system This often includes an ANM that manages, and in some cases contracts, the agents that provide cash and over-the-counter (OTC) transaction services to customers. 2 In each of the case studies presented in this paper, the funds are held by a risk sharing partner. However, under certain regulatory environments this role can also be contracted out to a bank (as in the case of M-Pesa in Kenya). For a detailed discussion on regulation of non-bank e-money issuers see “Non-bank E-Money Issuers: Regulatory Approaches to Protecting Customer Funds”, CGAP, 2010. Available at: http://www.cgap.org/sites/default/files/CGAP-Focus-Note-Nonbank-E-Money-Issuers-Regulatory-Approaches-to-Protecting-Customer-Funds-Jul-2010.pdf 8 PARTNERSHIPS IN MOBILE FINANCIAL SERVICES 2. The case study implementations Equity Bank Kenya: when banks and MNOs compete Since 2006, Equity Bank (“Equity”) has launched multiple mobile Money application which is marketed by Orange. Finally, Equity channels that involve bilateral partnership agreements with all but partnered with yuMobile to create yuCash, an exclusive e-wallet one of the MNOs in Kenya. The mobile and agent channels are the account for yuMobile customers. latest component in Equity’s very successful business model as a These partnerships have not evolved into robust MFS mass market financial service provider. With 7.1 million customers implementations. For Equity and M-Pesa, the M-Kesho product (June 2012), Equity is the largest bank in Kenya (measured by was a bridge between their respective core businesses. The two number of deposit accounts), with growing operations in the companies struggled to define the partnership in a way that satisfied region. Equity’s network of 165 branches has provided the basic their respective business interests, largely because they perceived infrastructure for Equity’s reach into the previously unbanked each other as competitors in too many parts of the business. Both and underbanked population. As early as 2003, Equity began companies have focused their efforts on their respective channels developing technology-assisted channels to expand its reach and neither company has promoted the M-Kesho product after the beyond branches. The rapid expansion of the ATM network in initial efforts at launch. In contrast, Equity and the other MNOs 2004 gave Equity a first mover advantage in geographic areas that speak positively about the contractual aspects of their partnership had been largely ignored by other retail banks. In late 2006 Equity arrangements. But these implementations have languished because also began developing mobile phone channels to further extend the MNOs are small relative to Safaricom and they have not made service delivery beyond the branch and ATM network. progress against M-Pesa’s dominant role in the market. The anchor of Equity’s mobile channel strategy is the bank’s Equity’s options changed in 2011 when the Kenyan Central own Eazzy247 service. Equity launched the service in 2007 to Bank issued the agent banking guidelines. This enabled banks to provide bank clients with a USSD-based mobile application for contract with agents to provide cash point services. Equity had transactional services linked to their bank accounts, around the believed for some time that it needed an agent network to fully same time that Safaricom launched M-Pesa. Crucially, in 2007 leverage its mobile channel and opted to build that network in- house rather than rely on the agent networks of others. As of June banking regulations did not allow banks to operate through third 2012, 1.78 million Equity customers are subscribed to Eazzy247 party agents. In contrast, the informal permission that the Kenyan and 5,004 agents comprise the agent network. Equity contracts Central Bank granted Safaricom did allow for agents and Safaricom with all mobile operators so that clients can access Eazzy247 on quickly created an impressive M-Pesa agent network. The take-off any network. As of June 2012, Equity customers were conducting of M-Pesa demonstrated just how critical the agent network is for 25% of all bank transactions at the agents on the mobile channel creating customer value with a mobile channel. So in 2010, Equity platform.3 In addition to portfolio and savings growth, the entered into partnerships with three mobile operators in Kenya in mobile/agent channel has also generated significant cost savings part to access the one part of an MFS channel that Equity could for Equity, given the lower cost of the channel relative to other not create on its own: the agent network. bank transaction channels. Between May and August of 2010, Equity formed partnerships The key point is that, in some cases, providers might be better off with Safaricom, Orange and yuMobile to create parallel MFS going it alone, where there is appetite and ability to do so. Despite channels. Equity partnered with Safaricom to create M-Kesho, trying a range of partnerships, Equity’s success has ultimately which is a unique Equity bank account that is linked to a Safaricom come without partnerships, relying instead on purely contractual M-Pesa e-wallet. Equity also partnered with Orange to create arrangements in agent network management and access to the Orange Money, an Equity bank account linked to an Orange MNOs’ USSD channels. 3 Interestingly, Equity serves every 100,000 customers with 2.32 branches, which is far lower than the average for the Kenyan banking sector at 4.4 branches per 100,000. This suggests that Equity’s alternative channels are increasing the efficiency of the bank’s infrastructure. http://data.worldbank.org/indicator/FB.CBK.BRCH.P5 PARTNERSHIPS IN MOBILE FINANCIAL SERVICES 9 WING Cambodia: when the core business evolves WING Cambodia Limited was formed in 2008 as a 100% responsibility for customer accounts in case of a WING failure. fully owned subsidiary of Australia & New Zealand Banking The resulting relationship between ANZ Royal and WING was Corporation Limited (ANZ), which had entered Cambodia in 2004 aligned with WING’s original business model as a value-added in partnership with the Royal Group of Cambodia, a local business service to the bank’s corporate customers. ANZ Royal wanted to conglomerate. The local bank, known as ANZ Royal, focused build a low-value payroll service to attract large garment factories as primarily on large corporate clients. ANZ launched WING in corporate customers, and WING was developed for that purpose. 2008 as a third-party payment processor, with a strategy to attract However, over time WING grew in directions that moved it away more corporate clients by offering value-added payroll services for from the core business of the partner bank, as WING shifted employees of corporations with large numbers of workers. from originating low cost accounts to promoting agent assisted WING consists of a mobile e-wallet platform and a network of OTC transactions for remittances and payments. This put pressure over 950 proprietary agents. Agents operate mobile phones or on both partners as WING acquired a growing customer base that POS terminals that are linked into the WING platform. Customers was not directly relevant to ANZ Royal’s corporate focus, and have the option of signing up for a WING e-wallet account and ANZ Royal became cautious about the potential operational risks accessing the USSD-driven transaction application on their mobile in the OTC services. This situation cannot resolve completely phones, or conducting OTC4 transactions at the agents without under current regulation, which requires a bank to assume legal having a WING account. WING’s services include cash in/out responsibility for a channel that is creating value for the MNO and at agents, domestic P2P transfer, bill pay, mobile airtime top-up, money transfer industries. ATM access, B2B services, retail payments, online payments and ANZ recently sold WING to Interlogistics Ltd, a Singapore- payroll distribution. Transaction volumes are concentrated in based company that also owns Refresh Mobile (Cambodia), which domestic P2P transfers and airtime purchases. boasts a network of 8,500 POS machines and integration with WING is a service provider and individual customer accounts are all mobile operators. ANZ Royal remains the sponsoring bank. held on the WING platform. The global float is held by ANZ Royal, Refresh Mobile and WING have now combined their businesses, and is settled between WING and ANZ Royal, usually three times branded as WING, to form the largest payment aggregator in the per day. The WING platform is interconnected with ANZ Royal’s market, allowing agent-based top-up for every mobile operator in network to enable WING customers use of the WING ATM card the market and leverage of their respective agent networks. at any one of over 115 ANZ Royal ATMs in Cambodia. As of July One of the key takeaways from WING is the importance of 2013, WING was processing $88 million per month in domestic aligning objectives and expectations upfront (including the target remittances transactions alone, and reports that it is now profitable. market). WING also highlights the importance of allowing the The National Bank of Cambodia gave permission for WING partnership to evolve over time, as the MFS implementation to operate on the condition that ANZ Royal assumed ultimate adapts to the needs of the market. 4 “Over-the-counter” (OTC) transactions refer to transactions that do not require the customer to have an e-wallet or bank account. For example, a customer might pay a bill by simply giving the agent cash. 10 PARTNERSHIPS IN MOBILE FINANCIAL SERVICES Easypaisa in Pakistan: the long-term view In 2008, the State Bank of Pakistan issued regulations that accounts on the platform while a pooled account on the Tameer clearly positioned banks (including microfinance banks) as the core banking reflects total balances held in all individual wallets. responsible party for MFS implementations. However, the MNOs Easypaisa offers a range of services similar to WING (airtime top-up, had a competitive advantage embedded in their communication OTC bill pay and P2P transfers, mobile transactions, etc.). channels and vast distribution networks and they were motivated The bank-led Pakistan regulatory framework may be interpreted by competition pressures to develop additional revenue streams. as relegating the most motivated sector - initially the MNOs - to Consequently, to be able to enter the MFS space, four MNOs a secondary role in MFS implementations. However, by allowing purchased microfinance institutions, either fully or partially; MNOs to take majority shareholder positions in microfinance another MNO entered into an exclusive partnership with a mid- banks, the regulations have created an avenue for MNOs to sized bank. invest and maintain sufficient influence over the implementation.5 Telenor was the first to move, purchasing a 51% stake in Tameer As a microfinance bank, Tameer is also motivated to exploit Microfinance Bank (which obtained a branchless banking license) the partnership to expand its core business. However, Tameer and launching the jointly operated Easypaisa service in October has approached the integration cautiously. When Tameer began 2009. The shareholding relationship has facilitated a revenue operating as a microfinance bank it experienced high non- sharing agreement with long-term objectives and a management performing loans (NPLs). Retaining close relationships with structure that integrates both organizations, making it more an customers was part of the solution used to reduce NPLs from acquisition than a traditional partnership. However, there are a over 20% to approximately 0.5%. Tameer has plans to pilot the use number of reasons why this implementation is relevant from a of Easypaisa to facilitate loan repayments, loan disbursements and partnerships perspective. Firstly, Tameer and Telenor each have savings mobilization from its customers, but will move slowly to specific roles and responsibilities. As the business and these ensure that the quality of the loan portfolio is not compromised.6 defined roles evolve, so too does the revenue split. Secondly, the Ultimately Tameer also plans to leverage this system to target the regulations ultimately prevent the relationship from merely being vast majority of Pakistanis that are not currently banked. the acquisition of a branchless banking license. That is, Tameer will remain ultimately liable to State Bank of Pakistan, and will In the Tameer/Telenor partnership, the acquisition united two be responsible for performing certain roles. Lastly, like other mass market service companies who both stand to benefit from partnerships, Easypaisa has had to deal with the challenge of the MFS implementation. Unlike the WING case, the lead investor merging employees from two different cultures, with employees in the business, Telenor, had strategic and economic reasons for from both organizations playing critical roles. being committed to delivering MFS to the mass market. But the ability of each partner to address the mass market was uneven. To date, the network has 25,000 agents that are using a mobile Telenor had the financial muscle and distribution capability to account operated from a USSD menu linked to the e-wallet reach large numbers of clients that Tameer was still too small to platform. Customers have the option of opening an e-wallet reach, and Tameer had the regulatory advantage of its banking account using the USSD menu for remote transaction or they license. Given the circumstances, the acquisition made perfect may elect to perform OTC transactions at agents without having sense. How this plays out over time in the core businesses of to open an account. Easypaisa reports nearly 5 million unique each of the partners still remains to be seen, but what is clear customers per month between OTC and e-wallets, with the bulk is that the ownership dynamic creates a longer term perspective, of transaction volume concentrated in airtime purchases and given that Telenor will ultimately share in revenue that accrues to OTC payments and money transfers. Tameer. It also reduces some of the misalignment of incentives The e-wallet platform is linked to a pooled account on Tameer’s that appeared in the other cases by closely aligning the interests of core banking system. Customers open individual Easypaisa the two partners. 5 It is also important to note that the level of investment required is significantly lower than would be the case if MNOs were forced to invest in full scale commercial banks. 6 “MFIs and mobile banking: blurring the lines?”, CGAP, forthcoming 2013. PARTNERSHIPS IN MOBILE FINANCIAL SERVICES 11 MTN Ghana: when the leader is forced to follow MTN Mobile Money was launched in Ghana in July 2009 as a The regulation and its addendum were intended to promote a partnership between MTN, an MNO, and nine commercial banks.7 “many-to-many” interoperable system, which would allow for The service is branded exclusively under the MTN banner, with seamless transactional flows between banks and MNOs offering no real visibility for the partner banks. mobile financial services. In practice, the regulations deprive not The MTN implementation has around 4,800 agents8 and an just MNOs but also banks of exercising a robust leadership role, e-wallet platform primarily accessed via cell phones through a requiring banks to assume legal responsibility for the key roles SIM Toolkit (STK) application.9 It also offers POS and internet in the implementation, but undermining the banks’ commercial channels, though these are relatively more recent and currently incentives to play those roles by creating a free-rider problem, being expanded. There is presently no card channel. Products as investment by any one bank in developing MFS would offered include P2P money transfer, bill payment, airtime top-up, automatically benefit its competitors.11 And indeed the banks have, balance inquiry and mini statement. Cash out is currently only with few exceptions, been unwilling to play an active part in the done at agents, but ATM cash out functionality is being developed key roles they are assigned by the regulation. Nor have they been in partnership with Fidelity and Ecobank. At the end of 2012, willing to invest significantly in product development, sales or MTN reported processing close to $30 million per month. marketing, since one bank’s investment would generate benefits for Regulations have created complex challenges in the Ghanaian all competing banks. For MTN Mobile Money, this has slowed the market.10 MTN aspired to first mover status in response to deployment and forced MTN to assume de facto responsibility for competitive pressure in the mobile communications sector, while most functions in the supply chain and invest significant resources banks had not shown an interest in pursuing mass market customers. into developing an agent and customer base that it legally does However, the Bank of Ghana issued regulations that permitted not own. Coordination of efforts across multiple unmotivated only licensed financial institutions to launch MFS, holding them partners has also proven to be a challenge, resolved only through responsible for agent network management, customer acquisition strong leadership on the part of MTN. and individual account management. MNOs are prohibited from filling those functions directly and thus effectively from playing The main takeaway is that regulations should ensure that the most a lead role. MTN initiated its first discussions with a single bank motivated players are enabled to compete on a level playing field, that at least partly shared MTN’s interest in reaching mass market with regulations proportionate to the risk of the products/services customers. However, an amendment to the regulations was being offered. Regulators must also be careful not to impede issued mandating a minimum of three banks in any branchless incentives of motivated stakeholders to invest and innovate, for banking partnership. example, by prematurely forcing collaboration. 7 The regulations required that MTN partnered with at least three banks. The nine banks were CAL Bank, Ecobank, Fidelity Bank, GT Bank, Intercontinental Bank, Merchant Bank, Stanbic, UBA and Zenith Bank. Since then, MTN has added the Agricultural Development Bank as the tenth partner bank. 8 Note: MTN does not use the word agent but refers to its agents as merchants. 9 STK “allows for the service provider or bank to house the consumer’s mobile banking menu within the SIM card” (FinMark Trust, “Mobile Banking Technology Options”, 2007), and can provide for an improved customer experience. 10 See e.g. http://www.cgap.org/blog/unintentional-consequences-branchless-banking-ghana 11 Many-to-many means that partner banks will be able to benefit from the investment of any one bank (for example in a new channel or product). 12 PARTNERSHIPS IN MOBILE FINANCIAL SERVICES 3. Aligning competitive forces, economic motivation and roles The starting point for understanding the economic motivations of the companies engaged in most MFS implementations is an understanding of the competitive forces in the banking, MNO, payment services, and agent- based distribution sectors. MFS channels create value by serving mass market customers and therefore the motivation of the players is driven largely by how much value they perceive that mass market customers can add in their respective industries. The case study implementations demonstrate that the most looking for a niche, and microfinance institutions with limited compelling business case may emerge from any of these sectors, branch infrastructure, to exploit an MFS channel to expand their and that the factors that give one sector more appetite than the mass-market business at relatively low cost. As noted previously, others will always be market specific. The case studies also reveal Equity grew to prominence in Kenya as a mass market bank and that the motivation can shift as an MFS implementation evolves. the addition of the agent network to its own Eazzy247 mobile Most MNOs are successful precisely because they have developed channel in 2010 was a cost-effective move that was consistent with a service with broad appeal to the mass market. Most MNO sectors Equity’s long history of developing technology assisted channels. have been extremely competitive, characterized by client churn Equity had long ago outpaced other banks in the Kenyan and falling revenues from traditional voice services. Consequently, mass market. The compelling competition in 2010 came from MNOs have spearheaded many of the first generation MFS Safaricom’s M-Pesa service. Mass market customers showed their implementations as value-added services with the aspiration considerable appetite for mobile payment services with M-Pesa, to distinguish their brand, lower distribution costs and add and Equity responded by positioning itself to compete in this new revenue streams.12 and obviously important dimension of the financial service sector. Banks may also have compelling motivation to invest in an The WING case demonstrates how economic incentives can MFS channel. In most banking sectors, the leading banks have shift in response to market realities, revealing an underlying established business models anchored in corporate and high misalignment in the original partnership arrangement, which net worth individual services provided through the branch network. These banks may add mobile channels as an additional grows as the implementation evolves. Early in its life, WING convenience to existing customers and may be genuinely had to adapt rapidly to manage several external challenges, interested in intermediating the funds mobilized through an MFS which constrained its ability to provide payroll services to ANZ’s implementation, but few have to date been interested in actually corporate customers. As a consequence, WING evolved over reaching the mass market with banking services. This represents time into a different business model which ultimately had limited an opportunity for aspiring mass-market banks, young banks relevance to ANZ Royal’s core business. 12 This describes the market conditions that MTN faced in Ghana in 2009. At the time, MTN had 51% market share, substantially more than the closest competitor Tigo with 23%. Since that time, competition has intensified and there are now six operators in the market, with MTN dropping to a 46% market share. Tigo and Airtel have also launched mobile money services and at least one other MNO is planning to launch mobile money services in 2013. In contrast, the competition for mass market customers is significantly lower in the Ghanaian banking sector. The entire sector serves only 34% of the population (compared to 85% of the population using mobile phones) and established banks have anchored their business model in corporate and higher income individuals. As a result, MTN and other MNOs have been eager to distinguish themselves through value-added financial services but only a few banks have shown significant interest in delivering bank products to the target market. PARTNERSHIPS IN MOBILE FINANCIAL SERVICES 13 Key lessons Market forces within the different industries will largely determine It is equally important to understand the influence of regulation. what kind of company will be most motivated and able to drive The case implementations provide several examples of an MFS implementation. Competitive dynamics between the regulations that force a suboptimal distribution of roles in MFS industries will determine whether companies from different implementations: sectors realize a net gain from an MFS partnership. • Regulation can prohibit the most motivated promoters • Partnerships thrive when the competitive advantage from exercising adequate control over their initiative. of the partners is clearly delineated, and the best This can occur easily in markets where MNOs or PSPs are eager placed partner is able to drive the initiative. to promote an MFS implementation under their own brand, The Tameer-Telenor partnership is an example. Telenor has a but regulations require banks to assume legal responsibility of clear advantage with regard to customer base, agent distribution the individual customer accounts or services. This is clearly the network, communication network, marketing and financial case in Ghana, where MTN has been the driving promoter but strength. For its part, Tameer brings its microfinance service cannot own the e-wallet accounts or the agent network. license, banking, risk management and compliance expertise • The lead position may need to shift as the business to the partnership. While Tameer is legally responsible for the model evolves. business, Telenor drives most of the customer facing aspects Most successful MFS implementations have evolved of business. significantly over time. The WING case demonstrates • Where banks and MNOs perceive each other as that this evolution can change the core business model significantly enough that the relative interest of the partners competitors, any partnerships between the two are may shift. Interests can be renegotiated, purchased or sold. likely to be difficult. But regulations that confine ownership to banks may strain The partners will be understandably wary of each other’s partnership arrangements when the core business model is no long-term aspirations and they will likely not support and may longer compelling for banks. even undermine any strategies that associate customer value with the other partner’s brand. This is clearly the case with Safaricom and Equity in Kenya, where both are competing for mass market customers with a large appetite for mobile financial services. 14 PARTNERSHIPS IN MOBILE FINANCIAL SERVICES 4. The core revenue sources in an MFS supply chain and the distribution of revenue The ultimate test of an MFS partnership is the revenue that accrues to the partners. An MFS supply chain links at least four distinct core functions that can generate benefits from an MFS implementation. Partners’ success is a function of how well the MFS implementation enables the companies in the supply chain to capture core sources of revenue. The following subsections explore how partnership arrangements enable companies in the supply chain to leverage all of the possible revenue sources. 4.1 Payment services13 to leverage the MFS channel in service of the bank’s core business. For Equity, for example, the MFS channel simply This is the most visible part of an MFS implementation. It allows the bank to reach more customers with an already encompasses the brand, the channel infrastructure and all of the very profitable array of banking services, but at less cost. The revenue that results from the transactions conducted in the channel. bank derives benefits by increasing banking service revenues These are typically cash-in/cash-out and payment transactions and reducing costs; by comparison, any service revenue from (e.g. remittances, bill pay, payroll services, etc). Revenue is channel transactions is barely relevant, at least in the short generated primarily from transaction fees and commissions and medium term. The bank may decide to price the payment paid by companies that use the channel to sell services or distribute payments. services accordingly. The WING case demonstrates that PSPs may struggle to optimize the payment service business when it Key lessons is legally owned by a bank. WING did not have control to fully develop the payment service business and ANZ Royal found • The relative importance of this part of the business can itself legally responsible for a separate company that no longer vary depending on the core business of the company supported ANZ Royal’s core business. that is operating the payment service channel. • The payment service business can also be undermined For an independent PSP, the payment service business is if the partnership arrangement or regulations do not primary to the business model, as in the case of WING. For allow the implementation to generate revenue from either a bank or an MNO-led implementation, this may not be and is probably most often not the case. This is because banks customer account balances. and MNOs can typically leverage the MFS channel to generate This is typically due to regulations but some partners do have more revenues in their core businesses than they will net from discretion in how this revenue is distributed. Some regulators the payment service. explicitly prohibit e-money implementations from earning any interest on the float balance. This is true of the trust vehicle • It is important that the partner that drives the payment used to sequester M-Pesa’s float balance in Kenya. It is also service business has sufficient control to maximize its true in Ghana. Even though MTN brands and operates the benefit. service, and holds individual customer accounts on their This is critical for a business model like WING, in which the e-wallet platform, the regulator prohibits MTN from earning company depends exclusively on channel related revenue. interest on the global float deposit in banks on the grounds that However, it may be equally important for a bank that aspires the funds do not legally belong to MTN but to customers.14 13 The payment service business could be operated by a bank, MNO or third party PSP. It is important to note that this section is referring to the business line (costs and revenues) rather than the provider of the service. 14 Unfortunately the regulations also lack provisions that enable non-bank MFS providers to pay customers interest on the balances in their wallet. While some MNOs would like to pay interest, they are wary of doing so for fear of drawing the attention of the regulator by engaging in what might be considered a banking activity. PARTNERSHIPS IN MOBILE FINANCIAL SERVICES 15 4.2 Banking Banks can generate core business revenues from MFS the bank sufficient control, visibility, and revenue share to implementations in two ways. The first way is to intermediate the warrant the investment. Moreover, banks will typically want mobile account balances by lending the funds to traditional bank to integrate services across all of their delivery channels. customers. The second is by providing banking services directly The Equity case illustrates the rationale of a bank that places through the MFS channel. A bank may have different requirements primary interest on using the relatively low cost MFS channels of a partnership, depending on the bank’s relative interest in these to expand its core business. For full control of the channel and two revenue sources. ownership of the customers, such a bank may well opt out of partnership arrangements. Key lessons The Ghanaian case demonstrates the problems that banks may have with too little control of the channel.15 The Ghanaian • Banks must be positioned to intermediate the account banks have little visibility in the channel, which is branded balances of an MFS implementation. under MTN. The regulations in fact explicitly forbid agents Float intermediation is typically a primary motivation for banks from using bank related terminology in branding as well as to participate in MFS implementations. Normally, banks can from cross-selling or marketing any of the participating banks intermediate the float balance whether they operate a mobile core products and services. The banks also face the difficulty banking channel and accumulate account balances directly in of distinguishing themselves by products, because any other the bank, or they intermediate the global float of an e-money bank can copy a product and supply it through the same provider. However, the Ghana case illustrates that regulations channel. Finally, client accounts reside on the MNO platform; can undermine this core benefit for banks. The large number the banks only hold the float balance of clients associated with of banks in the MTN implementation potentially dilutes any the bank. Banks cannot see individual customer information one bank’s income from intermediation of account balances, unless it is provided by the MNO. This discourages any and allows MTN to determine how the float will be distributed. significant investment by the banks. More importantly, the banks are effectively replaceable at the • Banks (and PSPs) face a particular challenge in markets MNOs’ discretion (with 30 days’ notice) and thus stand to where their MFS implementation competes against lose any investments they do make in the business should the other MFS implementations that are owned by the MNO decide to exclude them from the partnership. While this MNOs that provide the communication channels. type of 30 day notice clause is fine for a contractual service Non-MNO MFS companies face a substantially uneven provider relationship, it is clearly problematic where the bank playing field in markets where regulators allow MNOs to is expected to be a partner investing in the business. house MFS implementations within the MNO and where competition regulation does not prohibit MNOs from • Banks may require significant control of the agent manipulating service levels and pricing of communication channel to deliver banking services to clients. channels to protect their MFS business. This creates a potential A motivated bank will expend considerable effort and conflict of interest for MNOs and source of frustration for investment to develop, promote and sell banking products non-MNO MFS providers, such as Equity. This problem will through the channel, but this will be undermined if the bank become more visible if regulators allow MNOs to become does not have visibility in the client-facing part of the value MFS providers without introducing appropriate rules for “fair chain. Accordingly, the partnership arrangement must give channel access”. 15 It is true that most of the partner banks in Ghana are not explicitly wanting to drive the business or invest more. However, this does explain the challenge that a bank like Fidelity Bank, which is actively targeting the mass market, faces. 16 PARTNERSHIPS IN MOBILE FINANCIAL SERVICES 4.3 Telecommunications Like banks, MNOs also have multiple ways to generate revenues • MNOs may need to control much of the MFS channel from association with an MFS implementation. MNOs stand to to generate core business revenue. generate revenues directly from the transactions that are conducted Many MNOs expect to generate most of the benefit from an on their communication platform, and generate net benefits in the MFS implementation from the core MNO business revenue MNO’s core business. streams, at least in the near to medium term. However, banks and PSPs tend to want to make their channel available on Key lessons multiple mobile networks, and this can undermine the MNOs’ ability to generate core revenues. Moreover, bank and PSP-led • For many MNOs, the primary motivation for implementations will typically charge the MNOs commissions participating in an MFS implementation is the value for airtime sales. Therefore, MNOs may have significantly less added to the core communication business. to gain from a bank or PSP-led MFS implementation. MFS channels have become efficient ways of distributing airtime, reducing the overall commissions paid by the The Telenor/Tameer partnership seems to have overcome MNO. In some implementations, the average revenue per some of these disadvantages for Telenor. Telenor controls user (ARPU) has increased as MFS users increase airtime enough parts of the supply chain (communication channel, consumption because the purchase becomes more convenient. agent network, and e-wallet platform) to be able to generate core Finally, where MFS implementations function uniquely on telecom business benefits. Moreover, their ownership stake in a specific mobile network, the MNO may also benefit from Tameer also guarantees them a share of core banking business a reduction in churn as MFS customers are less likely to benefits, should these become more significant over time. abandon the network they use to access the services. These 4.4 The agent network revenue opportunities are becoming increasingly important to MNOs as airtime revenues decline. The Pakistan market The agent network is one of the most important components of illustrates a trend that is visible in most markets in the world: an MFS implementation. It provides customers with convenient in six years the ARPU dropped from USD9.00 to USD2.50. access to cash services, and this alone generates a large portion Therefore, forward looking MNOs are eager to add services of the customer value proposition. This makes the agent network to their network that can generate additional revenue streams strategically important to an MFS implementation, in particular and maintain loyal customers. to the company that wants to ensure that customers associate the service with their brand. Cash agent network management is a • Where MNOs play no other role in the implementation, highly specialized business that requires capacity that banks, PSPs they tend to maximize their revenues from charging and even most MNOs typically do not have. From this perspective, communication fees. a partnership that includes a company with an agent network, In some markets, the fees have been high enough to or at least an ability to manage cash across existing distribution undermine the economics of the entire MFS supply chain by channels, would appear to be efficient and mutually beneficial. undermining the customer value proposition. On the other However, the strategic importance of the agent network to the extreme, where MNOs own the MFS implementation or MFS implementation brings other considerations to bear on this are able to generate other revenues, they have been able to decision as “ownership” of agents has in some instances translated subsidize the communication expenses. into control of the overall MFS offering. PARTNERSHIPS IN MOBILE FINANCIAL SERVICES 17 Key lessons • MNOs typically launch MFS implementations with are not sufficient in the early phase to support an agent the aspiration to leverage their existing airtime business. In these cases, other companies in the supply chain distributors into a cash agent network. must push sufficient funding to the agents, and may be Their distribution networks make MNOs attractive partners required to wait for a period of time before seeing benefits to companies that do not have agents or any experience in accrue to their own core businesses. managing them. This is the case, for example, with Telenor. • Revenue sharing agreements between MNOs and However, the logistics involved in managing cash across this banks tend to be more complex. network have been a challenge for many MNOs. The research revealed a range of approaches to sharing • PSPs may also be motivated to build out their own revenues in these partnerships. On one extreme, banks and agent networks, for strategic reasons. MNOs have agreed to simply keep their respective core WING serves as an example. WING’s agent network is business revenues and share the payment service revenues. the foundation of its business model. The majority share For example, the partnership may split all transaction-related of transaction volume comes from OTC transactions at revenue while the MNO keeps all airtime sales revenue and the agents. From this perspective, WING is primarily an the bank retains all income related to float intermediation and agent distribution company that uses mobile channels and banking services. On the other extreme are agreements in a transaction platform to facilitate OTC transactions on an which banks and MNOs agree to share part of their respective agent network. The eventual sale of WING to a distribution core business revenues with the partner(s). The Tameer- company also demonstrates that the enduring value of the Telenor relationship is an interesting variation in that it gives Telenor a share of the global benefit through its ownership company is anchored in the distribution capacity of the agent in Tameer, regardless of whether benefit accrues in Telenor network. In short, the agent network is what enabled WING or in Tameer. This allows for some flexibility in the revenue to transform the initial business concept into a viable company sharing negotiation and agreement. The two companies share with long-term value. revenues in proportion to expenses incurred. 4.5 Revenue distribution • Given that most MFS implementations evolve significantly as they mature, the most important All of the companies in an MFS supply chain have to generate observation about revenue distribution is that it net benefit from somewhere. As the MFS implementation evolves, probably has to change over time as the revenue the benefits and costs may shift disproportionately. Revenue sources evolve. distribution agreements are the mechanism for balancing the It is very common for revenues to be concentrated in airtime interests of all companies in the supply chain, and are one of the sales and OTC transactions early in the implementation; most important aspects of partnership arrangements. banking services, P2P transfers and merchant fees are only likely to grow as the implementation matures. Partnership Key lessons agreements will likely have to evolve to maintain alignment • Revenue sharing with agents is common practice. between partner roles and their economic motivations, and This has been necessary to ensure that agents receive sufficient ensure that partners are well positioned to fully exploit the income early in the implementation to maintain their active revenue opportunities. Tameer and Telenor provide a good participation. Agents are critical to the core value proposition example of this, with the revenue split being reviewed on a of an MFS implementation but typically transaction volumes regular basis, as the offering, market and regulations evolve. 18 PARTNERSHIPS IN MOBILE FINANCIAL SERVICES 5. Final observations The research revealed multiple ways that companies partner to create value in an MFS implementation. The case studies provide meaningful examples of enabling partnerships. However, the findings are equally revealing about conditions that can undermine partnership arrangements. In fact, one of the noteworthy findings in the background research for this paper was the limited number of successful partnerships in operation. A second dimension to partnership requirements is that most • Prohibitions on non-banks earning interest on e-money float companies will want to leverage the MFS channel to generate balance that is deposited in banks denies e-money issuers an revenues in their core business. It is not enough to divide partner important source of revenue. This deprives the value chain of roles by their respective competitive advantage or installed an important source of revenue and essentially transfers value capacities to implement the MFS channel; companies may have from customers to banks. If a non-bank provider does not strategic reasons for controlling aspects of the MFS channel that earn float interest, then they must charge customers more. are most important to their core business. Alignment of all of these interests is challenging and critical to the success of any MFS • Regulations that force collaboration at an early stage of implementation. industry development can easily undermine the ability of individual companies to generate sufficient benefit to warrant Regulations are often an underlying cause of suboptimal partnership arrangements. A full examination of the regulatory the initial investment required in an MFS implementation. regimes in the four countries exceeds the scope of this exercise, but • Where banks and MNOs compete against each other in this finding warrants a brief summary of the specific regulations the MFS sector, MNOs will have a significant and unfair that can undermine the ability of companies to define roles that advantage unless competition regulation ensures that all make for effective MFS implementations. MFS implementations have equal access to communication • Regulations that require banks to assume legal responsibility channels, i.e. at a fair price. for all MFS implementations undermine the incentive and In conclusion, while many players are moving in the direction capacity of potential non-bank promoters. In many markets, non-bank players will have more motivation and competitive of partnerships for the delivery of MFS, these partnerships can advantage than banks, and bank-based regulation can deny take many forms and are inherently complex and dynamic. While these companies from leading an MFS implementation. there is no easy checklist for a successful partnership arrangement, this paper highlights many of the hazards that have affected early • Regulations that deny either banks or non-banks from partnerships – both in the commercial and regulatory arenas. owning and/or managing the agent network can easily deny What does seem clear is that, for a partnership to be successful, the leading company the control it requires over the primary source of customer value in an MFS implementation. Given each partner needs to enter into it with a clear understanding of the importance of agents as the primary face to clients, this its motivation, role and expectations, particularly regarding the creates a quasi-monopolistic situation that favors one sector composition and timing of benefits likely to flow to it. A long- over another, regardless of their competitive advantage or term view is almost certainly required, as is some degree of their motivation to lead an MFS implementation. flexibility in managing partnership relations over time. PARTNERSHIPS IN MOBILE FINANCIAL SERVICES 19 AUTHORS MARK FLAMING is a development finance specialist with over is currently working on topics such as the innovative use of data 25 years of international experience in developing financial to improve the adoption of mobile payments, microfinance institutions, regulatory frameworks and funding market institutions and mobile banking, financial inclusion partnerships instruments. He is a consultant with IFC and president of and interoperability in retail payments. He holds a Master’s Frontier Ventures, providing advisory services to financial service degree in Economics from the University of Witwatersrand. providers, regulatory authorities, socially responsible investors and development agencies working in developing economies. PETER ZETTERLI is the representative of CGAP’s Technology He holds a Master’s degree in International Development from and Business Model Innovation Team in Ghana and Nigeria, Harvard University. where he leads the work on understanding and catalyzing the development of branchless banking in the region. Prior to joining AIAZE MITHA is the founder of Amarante Consulting and a CGAP he spent seven years with the United Nations promoting consultant with IFC. Over the last 7 years he has advised mobile financial inclusion and private sector development in Sierra operators, financial institutions and policymakers on their Leone and China, where he tried to jumpstart Chinese mobile mobile financial services initiatives, providing strategic guidance, banking with Ericsson as early as 2008. He holds a Master’s regulatory advice, product development and implementation degree in Development Economics from Lund University. support across Africa, Asia and Latin America. He previously directed M-Paisa for mobile operator Roshan in Afghanistan. He GRETA BULL is the Program Manager for The Partnership for holds a Master’s of Science degree from University of Quebec Financial Inclusion and also manages IFC’s micro-retail advisory and a BA from Telcom Paris. services programs in Sub-Saharan Africa, covering microfinance, mobile financial services and insurance. She has 23 years of MICHEL HANOUCH works in CGAP’s Technology and Business experience in international development, half of which has been Model Innovation team, where he leads work on business spent in development finance. She has previously worked in models and the business case for branchless banking. He also Eastern Europe, the former Soviet Union and Latin America. She leads CGAP’s work on branchless banking in South Africa, and holds a Master’s Degree in Public Policy from Harvard University. 20 PARTNERSHIPS IN MOBILE FINANCIAL SERVICES THE PARTNERSHIP FOR FINANCIAL INCLUSION The Partnership for Financial Inclusion aims to bring financial services to 5.3 million previously unbanked people in Sub-Saharan Africa by 2017, by supporting the scaling up of sustainable microfinance and the development of mobile financial services. It is a $37.4 million initiative by The MasterCard Foundation and IFC that brings together the intellectual and financial capital of the Foundation with IFC’s market knowledge, expertise and client base. The partnership is also supported by donor and knowledge partners such as The Development Bank of Austria (OeEB, Oesterreichische Entwicklungsbank AG), the World Bank and CGAP. An important objective of the partnership is to contribute to the global community of practice on financial inclusion, and to share research and lessons learnt. This publication is part of a series of reports published by the program. To find out more, please visit www.ifc.org/financialinclusionafrica CONTACT INFORMATION Anna Koblanck Communications Officer 14 Fricker Road, Illovo 2196 PO Box 41283, Craighall 2024 Johannesburg, South Africa Tel: +27 11 731 3078 akoblanck@ifc.org August 2013