BLOG

Pitfalls in MFI Digitization: Getting Bogged Down in Technology

Look closely at recent innovations in mass market finance, and you will usually see some impressive new technology. What you will not see is how companies turned that technology into good digital products. And, if you’re a microfinance institution (MFI) looking at the market and thinking about offering digital products of your own, what you cannot see can hurt you.

Most financial institutions, including MFIs, will acquire new technology to take digital products and services to scale. For example, companies like Tymebank are demonstrating how fast and far mass market banks can move with a modern tech stack — specifically, a cloud-based, API-driven microservice architecture. But large technology changes are very challenging for banks, especially MFIs, to implement. Rushing into technology solutions, without having the institutional capability in place to commercialize them, is a common mistake MFIs make when digitizing.  As some companies have learned the hard way, new technology poorly implemented can easily destroy more value than it creates.

Using minimum viable products (MVPs) for early-stage technology innovation is the best way for most MFIs to test and iterate digital solutions, invest in building institutional capability as solutions prove themselves, and set themselves up for success.

So what does it take to create good digital products?

Business colleagues review information on a computer screen.
Photo: Arne Hoel/ The World Bank

Any change to an MFI’s products or operational processes needs to create value — both for customers and company. This is precisely where many expensive and difficult technology initiatives fail. Very often, companies use up significant resources attempting to implement a complex technology solution only to find that the value created does not compensate for the effort. In my view, based on experience in banks, MFIs and mobile network operators, the main reason for this is that companies often attempt to implement technology that they do not have the management culture, IT skills or product development experience to monetize.

One of the most important — and often overlooked — aspects of creating good digital products is having the right skills on the product team. Product owners need to lead product teams from development to market. To do this successfully with digital products, they need to develop the IT knowledge and skills to lead the product design and tech development processes through an agile practice. Technology implementations that are complicated or heavily dependent on vendor support make it difficult for product teams to lead the process. The complications tend to pull decision-making leadership up to the C-suite, back into the IT department, or even out to the vendor. This undermines the product team's ability to iterate rapidly through agile product design decisions.

An MFI’s management culture often needs to shift as well. Senior management needs to support the product owners as the primary decision makers related to product development and commercialization. Putting product owners in charge means that staff in other departments of the bank, in particular the IT, risk and compliance departments, need to work in the product teams. This shift is particularly challenging for C-Suite managers accustomed to dictating conditions that the business teams have to follow. The challenge is biggest for the CEO, who needs to manage this change to a more agile management culture.

It is tempting for an MFI management team to believe that it can acquire technology and gradually develop the required skills and adapt their management practices. But the learning cycle of typical technology implementations is too long. Companies that try to implement technology before they have the necessary management culture, IT skills and product development practices in place often experience costly, time-consuming failures that generate divisive frustration in the company.

The solution to this is to reduce technological complexity, focus on the product development practice, and shorten the learning cycle. This MVP approach makes it easier for companies to cultivate the necessary skills to create value with technology. Importantly, the MVP process also reveals to senior management the big institutional or personnel changes required for more ambitious implementations.

The MVP approach in automated loan origination

Thumbnail of report cover,
What can be learned from microfinance institutions that have digitized successfully? Click on the image to view this recent report.

Let’s use a concrete example to explore this idea. Many MFIs are interested in automated loan origination. One option that these MFIs have is to engage a third-party firm to build a credit scoring model, decision engine and customer-facing app. Development costs, revenue sharing and timeframe are often significant when going this route. Often, an MFI will task its IT department with leading interaction with the third party to manage the integration, security and data transfer aspects of the loan origination platform. This approach can easily take 9 to 12 months to produce an initial version. Often, that version is simply a digital replica of existing business rules and process flows, now embedded in a platform controlled by a vendor that charges for subsequent changes. Most importantly, this approach often preempts the early-stage customer-facing product development work that is more important to the eventual commercial success of this implementation.

There is an alternative approach that focuses more on early-stage product development and validation. In this approach, an MFI’s product team tests an MVP with clients to validate customer acceptance, using very simple technology. In this phase, the marketing, branding, customer interfaces, and terms and conditions can all be refined. The process flows are also worked out as the company transitions from the traditional decentralized credit officer-based loan origination process to a centralized, automated flow. Many operational details are sorted out. The risk department weighs in on all of these considerations. All of this can happen while the new loan decisioning model is executed manually on a spreadsheet and loan disbursements are batch processed manually.

In my experience, financial institutions gain a few important advantages with this second approach. First, it validates the key design decisions that drive customer and business value before making a significant investment. Second, it produces tested product and process flow specifications for the build out of the automated platform. And finally, it will reveal the capacity of the different parts of the MFI to work in an agile product development practice.

Most financial institutions will have to make significant changes to their skill sets and management culture to build value with the emerging technology. Companies that deploy MVPs and focus their early-stage efforts to build these capabilities will build value faster than companies that adopt complex technology before they are ready.


Getting bogged down by technology is a common pitfall in MFI digitization, but it is not the only one. To learn about other common yet avoidable issues, see the other posts in our ongoing blog series, "Pitfalls in MFI Digitization: What They Are and How to Avoid Them." Also see our publication, “Digitization in Microfinance: Case Studies of Pathways to Success,” which offers an in-depth look at the journeys of several MFIs that have digitized successfully.

Sub-topics: MFIs

Add new comment

CAPTCHA