Technology can be an instrumental tool in the scalability of a microfinance institution. Most organizations realize early on that they need a strong management information system (MIS), but many struggle to find one that can be properly implemented and operated by staff in challenging work environments.
Have we Made Progress?
We all know that there is no “gold standard” for back office software. If microfinance was a simple and easily replicable process of giving a loan and getting it back, why wasn't there one leading microfinance software provider? Indeed, there were many organizations that claimed to have solved back office and front end issues, but upon closer inspection one can find out that more often than not, initiatives started with good intentions but fell short of the hype.
The main problems comes in two categories unrelated to technology: one, seemingly effective sales and service models designed on a global scale and localized for dozens of languages; and two, platforms customized to the organization's unique operational methodologies. The third issue pertained to infrastructure: technologies designed for office buildings in developed countries can rarely be applied to connectivity-challenged, energy-starved, dusty, hot, miles-from-any-technical-person-location
Today, the market is changing. We see a number of new efforts underway, from Mercy Corps starting afresh with new software via a US$19 million grant from the Bill and Melinda Gates Foundation to Crane Software selling Banker's Realm across East Africa; from open source programs like Mifos and Octopus to several mainstream companies launching software as a service model on a global, regional or country-wide basis, e.g., IBM and Salesforce. Fern Software and Southern Horizon, who each had a small but meaningful part of the market, recently combined their product lines.
As an MFI manager, what would I make of these developments? What should we, as a sector, make of this? Are we making progress?
Technology is a Tool
The chief question asked by microfinance leaders is not, “Should I use technology?”, but rather, “What is the best technology to use?” Or, “Why is my technology project stuck?” And with time the question has become, “How can I replace this drag on my organization's efficiency?”
It is not merely in microfinance that we find general unhappiness with information technology, for there are many studies that show that technology projects have, on average, quite a poor record of perceived success. As a technologist, we must admit that we have ourselves to blame, at least in part, for over-hyping the technology and under-emphasizing change-management issues.
MFI managers is to treat information systems as a strategic strength or, in the absence of one, as a threat. If you don't have at least a workable system, you have a problem; the earlier you recognize this, the better. If you intend to grow, you will need to provide more and more data to funding agencies and generally become more capable of identifying operational and market risks than you were in the past. This requires “all hands on deck,” as well as solid management thinking, and cannot be assigned to the guys that fix your PC hardware.
Unfortunately, even when they do all the right planning and analysis, MFIs often realize that either they cannot afford a system or that the systems available won't meet their requirements. This shouldn’t be surprising. This fundamental problem relates to a typical tension in software systems design: using “enterprise systems” versus commercial off-the-shelf systems (COTS). Enterprise system design, which is a more expensive and tailored approach, requires a close alignment of system logic to the existing procedures of the MFI, while off-the-shelf systems generalize products and workflow, and utilize variable parameters.
Most of the systems available for microfinance try to split the difference, often resulting in costly complexity and unhappy users. Put another way, users are generally happy with software that does what it claims to do—a spreadsheet program—and generally unhappy with code that attempts to match all areas of functionality.
While a few MFIs have managed to build their own systems internally, the history of information technology is against them, and eventually market pressures will push them out of the business of building their own.
On the Horizon
Mifos attempted to solve several of these dynamics by giving some amount of flexibility to the MFI, while retaining a set of core operational and transactional aspects for a common platform. Unfortunately, for many reasons, the Mifos value proposition has failed to materialize (as of yet) as a driving force for locally-driven software evolution. Until a significant number of global vendors leverage a solid, open Mifos core that is profitable and provides a highly scalable system to MFIs, the critical mass necessary for an open source model cannot be developed. It may be that such a model is not possible given the competitive nature of MFIs and vendors, but to date this has not been proven either way. Perhaps Grameen Foundation is right in privatizing the ongoing development in a for-profit model, but without more transparency the industry will lose an opportunity to explore a public-good approach to what is certainly a basic infrastructure need for microfinance operations: transaction tracking.
Filed under the category of “Didn't Expect to See That” is the news that MFIs are looking more seriously at big ticket banking software, and most significantly, banking software providers are looking back. As MFIs grow and become more aligned with banking standards, and banking software providers look for new markets, there is more convergence in features available and relative affordability. This is good news for those players large enough to pay US$1m (or more) for a system, although much less useful for many of the small and medium sized MFIs out there.
Shared Plaftorms, a type of “outsourcing,” has received attention lately, and takes a more streamlined approach, essentially providing a lower cost way for MFIs to handle a set of operations in a typical fashion, but without the in-depth level of business logic customization found in an enterprise solution. These shared platforms, whether at a regional or country level, require stable online access and thus, at this time, fail the general filter for MFI managers, who see “always on” connectivity as unrealistic in their head offices, let alone the branch offices.
There are hybrid approaches with a certain amount of offline capabilities, but fundamentally computing moves off site. Salesforce has an “on-demand” model that allows users to pay only for the number of users needed, and has a microfinance edition with some basic features for microfinance on top of their solid platform. Microbanx tried this model several years ago and it didn't take off in a big way—but timing is everything, and perhaps now connectivity has reached an important inflection point.
This brings us to mobile networks and the ubiquitous data connections they provide. As we enter an age where the mobile phone is used by more than any other information communication technology that has come before, we should contemplate just how remarkable it is to see finance for the masses converge with ICT (Information and Communications Technology). While we have seen plenty of hype around this as well, we are also seeing a growing set of players who are building single-purpose transaction models on the mobile phone. These include a surprising market for ringtone purchases, top-off value transfers, and remittances. These are models that actually work and earn good returns for investors today.
Can microfinance for the masses be executed on a largely impersonal basis via phone or is the loan officer an essential element? Can MFIs forgo the branch model entirely? Will we see a few large global MFIs serving most of the BOP with banking systems paired with mobile technologies? Or a plethora of small players creating endless new models in new markets with new services?
Technology enables us to ask these questions. These are exciting times.