Why Ethiopian microfinance must build digital readiness before making digital commitments
What concerns me is not that Ethiopian microfinance institutions are moving too slowly into digital finance. My concern is that some of us may be trying to become digital before we are digitally ready.
I began the African Microfinance Weekly Media Digest nearly six months ago for a simple reason. I was already following developments in microfinance, financial inclusion and technology for my own learning, and I felt those developments were worth sharing with others in the profession. As the weeks passed, however, the reading began to reveal something more than news. One concern kept returning. Technology is moving quickly, while many of the institutions expected to make decisions about it still lack the knowledge, systems and governance needed to lead those decisions.
I am not an information technology expert, and I do not write as one. I write as a business development practitioner who has spent the past five years working within the microfinance sector, with opportunities to engage closely with institutional leaders, regulators, technology providers and development partners. I have also sat inside discussions where major technology decisions were being considered and watched what happens when technical knowledge, management confidence and governance capacity are badly unbalanced. I have seen how easily a product can be discussed before the problem is clearly defined, how a provider can control the conversation because it controls the language, and how easily an institution can become too dependent on one technical specialist when others either lack the knowledge or are not given enough time, space and authority to question or independently verify the advice being given.
These experiences have left me with a difficult question. Are Ethiopian MFIs building digital capability, or are we acquiring the appearance of being digital?
The appearance of transformation
Digital transformation is increasingly confused with acquiring visible signs of technology. A mobile wallet, an agent banking channel, a mobile application or a fintech partnership can easily be presented as evidence that an institution is becoming modern. These are visible developments. They are easy to mention in a board report, a social media post or a speech, and they can create a strong sense that the institution is moving forward.
But visibility is not transformation. A new channel is not a digital strategy, a signed agreement is not an institutional capability, and a software licence is not a transformed business.
Real digital readiness is more basic than launching a visible customer facing product. An MFI may have a core banking system and still conduct much of its daily work manually. Approvals may depend on paper, staff records may sit outside integrated systems, routine workflows may remain unautomated, and employees may rely on outdated computers that slow even ordinary tasks. In that environment, adding a wallet, an agent banking channel or a mobile application does not transform the institution. It simply places a digital layer over operations that are still largely manual.
I have observed discussions in which a wallet, agent banking, a mobile application and other digital channels were already being treated as priorities before the institution had clearly defined the problems they were meant to solve. Different products with different purposes, operational requirements and regulatory implications could be discussed almost as one digital package. Who needed each service, what problem it would solve, whether the institution needed to build something new at all, and whether it was ready to manage it had not been properly tested. Yet attention had already moved to providers, approvals and launch dates.
That is not digital transformation. It is digital ambition without institutional preparation.
The first act of digital leadership should therefore be simple. The institution should begin by deciding what it actually wants to improve. It may be faster loan collection, easier deposits and withdrawals, lower cost outreach, better information for credit decisions or less cash in the hands of field officers. Different problems may require different solutions. Starting with the technology instead of the problem risks spending scarce resources on something the institution never truly needed.
When the seller also defines the problem
The second concern is the knowledge imbalance between many MFIs and the companies approaching them.
A fintech or technology vendor normally arrives prepared. It understands its product, the architecture, the commercial model and the language of the negotiation. It has a presentation, a demonstration, financial projections and answers to expected questions. It may have negotiated similar arrangements before. The MFI, on the other hand, may be entering this kind of discussion for the first time. Its board and management may not have enough independent knowledge to test what they are hearing.
This creates an unhealthy situation. The company selling the solution can end up defining the problem, designing the solution, estimating the opportunity, valuing what each party contributes and explaining why the agreement is attractive. The MFI appears to be making a decision, but most of the decision has already been framed by the other side.
The issue is not that technology companies are automatically dishonest. Many fintechs have valuable knowledge and can help MFIs serve clients better. But every provider also has its own commercial interest and will naturally present its solution in the strongest possible light. The difficulty begins when the MFI does not have enough independent knowledge to test that case properly. In that situation, the provider’s understanding of the problem, the solution and even the value of the deal can quietly become the basis on which the MFI makes its decision.
Consider a profit sharing proposal. A fifty fifty split may look balanced, and a sixty forty split may appear even more attractive to the MFI. But neither percentage tells the institution whether the deal is fair. The real question is how that share was arrived at. The MFI may contribute the lending capital, cost of funds, customer base, field presence, collections capacity, regulatory responsibility and years of trust in the market. The fintech may contribute the platform, product design, technical operation and other specialised capabilities. Each side may also carry different costs and risks. Until those contributions, costs and risks are properly understood and valued, a percentage is only a number.
The institution also needs to understand which costs are deducted before profit is calculated, and who controls pricing, credit assessment, loan approval, collections and customer communication. It must also understand how financial, operational, regulatory, customer and reputational risks are divided between the parties. An attractive commercial term should never substitute for understanding the full allocation of risk and responsibility.
If these questions are not answered in language that the board and management can understand without the vendor in the room, the MFI is not ready to sign.
When one person becomes the strategy
The stories I have followed through the Media Digest show how quickly digital finance is expanding around African microfinance. New platforms, partnerships and technology driven models are becoming difficult for MFIs to ignore. But the pressure to move quickly creates another risk inside the institution. When a specialist is brought in to lead a major digital change, and management and the board are not equipped to properly question the direction being proposed, that person can gradually become more powerful than the governance system around the change, even when the specialist is capable and acting in good faith.
Hiring an experienced IT leader may be necessary, especially where specialised skills are scarce. But hiring one strong person is not the same as building a strong IT function. Neither is changing the organisational structure. If the wider team remains poorly equipped, insufficiently trained and unable to question or carry major work independently, knowledge simply becomes concentrated at the top. The institution has not solved its capability gap. It has concentrated it.
That dependence becomes more serious once major decisions and investments begin to accumulate. Systems may be changed, contracts signed, data migrated and new services placed on the roadmap. The same person may then become the main source of explanation for what has been done, what remains unfinished and what should happen next. Losing that person can begin to look like a risk to the entire programme. Management may then become reluctant to challenge decisions it cannot independently verify because the cost of losing the specialist appears greater than the cost of continuing the dependency. At that point, this is no longer an ordinary staffing issue. It is a governance risk.
The problem becomes worse when legitimate questions are treated as resistance to change. Someone who asks whether a product is really needed, whether the institution is ready, whether the cost is justified or whether another approach should be considered can easily be seen as protecting old ways of working or standing in the way of progress. That is a dangerous confusion. Questioning how transformation is being done is not the same as opposing transformation. A change programme becomes difficult to govern when challenging the approach is treated as rejecting the goal itself.
The answer is not weaker IT leadership. Ethiopian MFIs need stronger technology professionals, stronger IT teams and much greater investment in specialised skills. But knowledge must be shared, staff must be developed, major decisions must be documented, and management must have access to more than one informed view. Business, operations, finance, risk, compliance and internal audit must also have enough understanding and authority to challenge decisions from their own areas of responsibility. Strong technical leadership should make the institution more capable, not more dependent on the person leading it.
The test is simple. If an MFI cannot confidently explain, question or continue its digital direction without one person in the room, it has not yet built digital capacity. It has built digital dependence.
A shared system is not a sign of weakness
The question of core banking deserves particular care. Some MFIs may believe that leaving a shared system and acquiring a system directly is proof of greater independence or digital maturity. This is not necessarily true.
A shared core banking arrangement can be a sensible choice for smaller and medium sized MFIs when the service is reliable, secure, responsive and well governed. It can spread costs across institutions, provide access to specialised technical capacity, support common standards and strengthen bargaining power with technology providers. But shared systems can also create problems when responsibilities are unclear, service quality is weak or participating institutions have little influence. The arrangement should therefore be judged by how well it works and how well it fits the institution, not by whether it appears more or less independent.
Shared core banking models should not be dismissed as temporary arrangements for institutions that lack ambition or digital maturity. They can be deliberately designed to automate core processes, respond to different institutional needs and allow future integration with fintechs and other digital service providers. When properly structured, they can also bring together planning, system selection, testing, configuration, migration and technical support that individual MFIs may struggle to manage alone. The value of the model should therefore be judged by its capability, governance and service quality, not by the fact that the system is shared.
Whether a shared arrangement is actually serving each institution well is a separate question and should be judged honestly. Poor service, weak support or limited flexibility are valid reasons for an MFI to consider leaving the shared arrangement. But acquiring the same or a similar system independently does not automatically create greater digital maturity. It may simply transfer more cost, implementation work, security exposure and technical responsibility to one institution. If that institution is not ready to manage those responsibilities, greater ownership of the system can actually create greater dependence.
Ownership of a copy of software is not ownership of digital capability. Before leaving a shared arrangement, an MFI should be able to show what will genuinely improve for the institution and its clients. It should understand the full cost, the demands of migration and integration, the capacity required to manage the system, and how operations will continue if implementation fails or is delayed. Independence is valuable only when the institution has the capacity to carry the responsibilities that come with it.
What the MFI may slowly give away
An MFI brings assets to a digital partnership that do not always appear clearly in a financial model. It brings its licence, capital, customers, community trust, local knowledge, field presence, collection experience and access to a market that may have taken decades to build. A fintech brings different strengths, including technology, speed, specialised knowledge, product capability and a better digital customer experience. A good partnership can combine these strengths. But before discussing how the benefits will be shared, the MFI must first understand exactly what kind of partnership it is entering.
Not every fintech partnership is the same. In one arrangement, the MFI may simply purchase technology while retaining control of the product and customer relationship. In another, both parties may jointly distribute and manage a service. In another, the fintech may control the customer interface and much of the product experience while the MFI provides the regulated lending capacity behind it. These arrangements create very different positions for the MFI. The danger begins when an institution enters one model while believing it has entered another.
The balance can also shift gradually after the partnership begins. If the fintech controls the interface, the data, the customer journey, the credit model and communication with the client, the MFI may slowly lose the ability to understand and manage its own business. It may continue to provide capital and carry regulatory responsibility while another company learns from its customers, strengthens the relationship and builds the more valuable part of the business. The MFI can eventually become little more than the regulated balance sheet behind somebody else’s financial service.
This is why questions about the customer relationship, data access, product decisions, pricing, credit assessment, complaints, regulatory accountability, knowledge transfer and exit rights are not legal details to be settled after the commercial agreement. They are the business model itself. The MFI needs to know what it will continue to control, what it is willing to share and what capability will remain inside the institution if the partnership ends.
Collaboration is necessary. Most MFIs cannot and should not build every technology themselves. But a partnership should leave the institution better able to understand its customers, make decisions, manage risk and serve its mission. It should not make the MFI more dependent on the partner with every year that passes.
Regulation does not allow us to outsource responsibility
The most useful Ethiopian reference for this discussion is the National Bank of Ethiopia Directive MFI/33/2022 on information technology management for microfinance institutions. The directive does not tell an MFI to begin with a wallet or an application. It requires the institution to define the role of technology within its business strategy and to develop an IT strategy that supports that direction. It also places clear responsibilities on the board, senior management, the IT function, risk management and internal audit.
The requirements go much deeper than having a system in place. The directive calls for proper planning, adequate financial and human resources, project and vendor management, automation of core business processes, reliable management information, technology risk assessment, disaster recovery, training and IT audit. It also requires the board and senior management to review technology matters regularly. Taken together, these are not simply compliance requirements. They describe what institutional readiness should look like.
The wider national direction reinforces the same message. The draft National Digital Payments Strategy for 2026 to 2030 places human capacity, trust, resilience, supervision, consumer protection and shared infrastructure alongside innovation. It does not treat digital progress as a race to acquire more technology. It treats it as a system in which infrastructure, people, governance and responsible use must develop together.
The same principle applies when an MFI works with an outside technology provider. International guidance is clear that using a third party does not remove the institution’s responsibility. The MFI still has to understand the risk, assess the provider, define responsibilities clearly, monitor performance, prepare for failure and retain a practical way to exit. It must also keep enough knowledge inside the institution to manage the relationship itself. A provider can supply technology and expertise. It cannot replace the judgement that properly belongs to the institution.
What real digital readiness requires
The more I have followed digital developments through the Media Digest, the more one thing has become clear to me. An MFI is not digitally ready because it can buy a system, sign with a fintech or obtain approval for a new channel. It is digitally ready when it understands the problem before choosing the technology, understands the technology before committing to it, and understands the consequences before putting its clients and institution behind it. Being able to launch something is not the same as being ready for it.
Readiness begins inside the institution. If approvals still move on paper, records are difficult to retrieve, basic processes remain manual, reports cannot always be trusted and staff struggle with inadequate tools, a new digital product does not make those weaknesses disappear. It carries them into the new system. Technology can make a strong process faster, but it can also make a weak process fail faster and at a larger scale. Digital transformation cannot begin at the customer interface while the institution behind that interface remains largely unchanged.
It also begins with knowing what you want. Before asking which wallet, platform, application or provider to choose, the institution should be able to explain the problem in plain language. Whose problem are we solving? What is difficult for the client today? What will become easier tomorrow? Why does this particular technology solve it better than the alternatives? If those questions are difficult to answer without the vendor in the room, the institution is already moving too fast.
Real readiness is also visible in how confidently an institution can question a proposal. Management should understand enough to challenge the economics. The board should understand enough to challenge the direction. The IT team should be strong enough that knowledge does not sit with one person. Business, operations, finance, risk and audit should understand what the change means for their own responsibilities. An MFI does not need to know everything a technology company knows. But it must know enough to decide for itself. Otherwise, the institution may own the licence and provide the capital while somebody else effectively owns the thinking.
Readiness also means having the confidence to say no, or not yet. Digital transformation should not become a race to collect channels and products. An MFI can start with the need that matters most, pilot one solution, let staff and clients live with it, learn where it works and where it strains the institution, then improve and expand. The institution should feel in control of one step before rushing into the next. That is not moving slowly. It is building the capacity to move further without losing control.
For me, this is the real test of digital readiness. Can the MFI explain what it wants, choose deliberately, challenge what it is being told, support what it builds, learn from what happens and change direction when necessary? Can it do all of this without becoming dependent on one vendor, one system or one individual? If the answer is no, the institution may be moving toward digital finance, but it is not yet ready to lead its own digital transformation.
The greater danger is not being late
Ethiopian MFIs do need to change. Clients expect easier access, faster service and greater control over their money. Digital finance can help us reach further, operate better and serve people in ways the traditional model could not. Standing still is not an option.
But neither is moving simply because everyone else is moving. The greater danger may not be arriving late. It may be committing to systems, products and partnerships that the institution does not yet understand well enough to govern. An MFI is not weak because it needs a fintech partner, uses shared infrastructure or takes time to prepare. It becomes weak when it can no longer explain its own choices, challenge what it is being told or operate without those it has become dependent on.
Perhaps every major digital decision should face one simple test. If the system or partnership ended three years from now, would the MFI be left with stronger people, better knowledge, better data and greater ability to serve its clients? Or would it discover that much of the capability it thought it was building never truly became its own?
For me, this is what digital maturity ultimately means. Not having the newest technology, but becoming an institution that knows what it wants, understands what it is committing to and remains capable of making its own decisions. The MFI should understand the deal at least as well as the vendor. Until it does, it should not sign.
