Making microfinance viable with IT

from Network World

By Kanika Goswami

Kanakapura is a township of about 50,000 on the fringes of Bangalore. It's known more for the road that leads to it: Kanakapura Road, on which an Art of Living Center is located.

But it's also the hometown of Razya. For three years, the 28-year-old could not find work in the small town and her husband barely made ends meet with his job in the sericulture business.

That all changed when Grameen Koota, a microfinance institution, lent her enough money to start a scrap business. In its first year, the business earned the family a profit and Razya was able to send her kids to school and purchase some furniture.

Razya is one of many. Although most others are not as fortunate. It sounds unbelievable but Bangalore, arguably the world's IT capital, is the 13th poorest district in Karnataka. An estimated 38 percent of the district's 1.7 million people live below the poverty line.

Related Content

Experts agree that development among this class of poor is hard. Mainly because the cost of servicing microfinance loans (under Rs 30,000 (US$700) doesn't make business sense and commercial banks don't trust the poor to repay their loans.

That's a view that's beginning to change. "Microfinance is needed for people who are left out of formal financing institutions. Their inability to pay is never a reason, they pay back better than others," says Suresh K. Krishna, MD, Grameen Koota. According to Grameen Koota's website, it's repayment rate is 99.99 percent.

The high administration costs associated with small loans, however, is a harder nut to crack. Some experts estimate that servicing a microfinance loan is about six times more, proportionately, than servicing an average bank loan.

These problems kept commercial banks from seeing the poor as a potential market until Grameen Koota showed them.

Sorry, We Can't Help

When Grameen Koota started in 1999, its aim was to help poor women in rural areas and urban slums through micro credit. Their target audience were women who earned between Rs 40 to Rs 80 a day. If they could consistently deliver need-based financial services to these women -- in a cost-effective manner -- they could be a financially-sustainable microfinance institution to the poor.

The Grameen Foundation, which also works with the Grameen Bank in Bangladesh (which got the Nobel peace prize for its work in micro-credit), supplied Grameen Koota with Rs 44 lakh to start work in the Bangalore district.

Today, the institution's mission hasn't changed -- but its way of working has. When it started, it worked with a microfinance management application. Like all start-ups, it was satisfied with an application that met its basic needs: track its portfolio and accounting for microfinance.

But growth started catching up with the software. Maintenance issues cropped up, MIS reports took longer to create and expanding the institution's operations was becoming a problem.

"For example, [the system] could not maintain transaction data," says Krishna. It overwrote "each transaction on top of the last one, maintaining only balances." This meant that a loan officer in a remote village could not tell when the last repayment was made -- only how much more needed to be paid. Since transaction records were maintained off the system, "you had to do a lot of work to generate proper transaction data," Krishna says.

And the system's simplicity, once a measure of its success, began to be a drag on efficiency because it did not allow modifications to existing accounts. For example, if the term of a loan's repayment was to be changed from 60 weeks to 48 weeks, a loan officer had to create a new account. These tedious workarounds left Grameen Koota unable to handle the volumes it needed to meet its growth targets.

Related Content

Worse, the software was unable to support new product innovations. It could handle only three products: a general loan, a consumption loan and a supplementary loan. This locked Grameen Koota out of a world of other loan and insurance possibilities and hampered its growth.

Growth, however, was exactly what the institution wanted. After its initial success, the institution, which will soon get the status of a non-banking financial institution, wanted to spread the goodness. So it set itself some difficult targets. It's strategic goals for 2007-08, for instance, aims to increase both its client base and its portfolio by a 100 percent.

To do this, it needed to be able to open new offices quickly, be more efficient in the areas it already serviced, and create management reports much faster than it already did, so that decision makers could take calls like opening new branches.

It also needed a system that was flexible; since every branch might want to tweak its system. For example, how interest is calculated differs by region.

"The flexibility and scalability of the [new] product means that we'll be able to simultaneously standardize common processes, accommodate regional variations, and scale for new innovations in the future," says Krishna.

Spread the Goodness

In a bid to get a system that would meet all its current requirements and handle expansion, Krishna and his team walked through all the available products but failed to find a generic product that met their needs. They also realized that an off-the-shelf product would leave them more 'vendor dependent' than they were comfortable with -- especially since the expansion would require local customizations.

The option of developing a system inhouse was considered, but was quickly scratched out because it would require too much specialization.

It was around this time that the Grameen Koota team figured out that an Open Source solution would fix all their problems: it was scalable, it would free them from vendor-dependency and it could be customized at the branch level. And it would be less expensive than a bought solution -- an important criteria.

"Applying an Open Source model to a microfinance application is a very innovative approach to tackling the severe cost pressures in the microfinance sector. It helps MFIs (microfinance institutions) profitably extend their reach to poor communities," says Lee E. Tenny, managing consultant, IBM Global Business Services, Financial Services-Strategy & Change.

After it jotted down its goals, Grameen Koota identified a solution: Mifos, an Open Source management information system designed specifically for the microfinance industry. Mifos, developed at the Grameen Technology center at Seattle with help from IBM, allowed individual MFIs to modify the software to their needs. It also provided the key functions required by a MFI including client management, loan and savings portfolio management, loan repayment tracking, fee and savings transactions, and reporting.

For Grameen Koota, the journey to implement Mifos has been an arduous one, partially because the institution also played an important role in developing the solution further.

The challenges were on two fronts: migration and training. Legacy data needed to be rediscovered or corrected before it could be migrated -- from 44 centers spread across 13 districts -- onto Mifos. In the meantime, training manuals and infrastructure had to be created.

"Transforming all the data into the new format was really challenging. We had about 1.2 lakh deviations,' remembers Krishna. For example, there were loans in the old system that showed up as a late payment, but Grameen Koota said there were not. So, loan repayment schedules had to be regenerated. "We needed to migrate all the data, but since the transaction history [of a loan] was not part of the earlier application, we had to run an intermediate script and get all the data in place, then put it into the format required for Mifos. Though nothing was lost it, took quite a bit of time to migrate," he says.

The migration needed to be done branch by branch. For a while, each branch used both their existing system and Mifos in parallel. Each week, data was re-migrated to Mifos on the day when no meetings were held. This was done to capture any changes made in the legacy system during the week, since Grameen Koota didn't have enough staff to enter changes on both systems. This introduced performance issues because with tiny payments every week the database kept expanding.

In retrospect, some people associated with the project think it would have been easier if the migration could have been done with a merger tool and a separate migration tool.

Training everyone on the new software would also take away much mindspace. Training was done inhouse, but in a staggered manner, across the centers. "We took training sessions at many places, clustering some branches and training one set of people at a time. Over the five months that it took to migrate, half were trained, and then they trained the other half," Krishna says.

The team quickly set up a help desk to aid with issues during dry runs. Its aim was to identify operational issues and rectify them. Documentation of all processes was also carried out in this phase. Since Mifos is an Open Source application, every step needed to be documented and updated on the Mifos site, so that other users could benefit from new customizations. To play safe, Krishna hasn't yet pulled the plug on the old system. "I have not removed our legacy software even now. So far, nothing has happened and I plan to go ahead with only Mifos now," he says.

What helped, everyone agrees, is that the ground was ripe for user acceptance and users and the IT team were enthusiastic.

Link to full article. May expire in future.