Credit and Fraud: How One M-PESA User Weighs Risk on Kenya's Largest Mobile Money Platform

8 minute read

I was recently in Kisumu, a semi-rural area on the coast of Lake Victoria in Western Kenya. I was there for ten days; training organizations on our mobile payments software, collecting indicators to monitor our DFID project and connecting with organizations to work with in the future. One of the places I spent the most time over the week was Support for Tropical Initiatives in Poverty Alleviation (STIPA), which is an amazing organization with projects in microfinance, health insurance, and economic opportunity programs. My time at STIPA was made very worthwhile by the dedicated, knowledgeable and passionate staff. We’ll be working closely with STIPA as they merge two of their major programs, Community Based Health Financing and Village Savings and Loaning to begin using mobile payments. They’re two very impactful projects with unique challenges and potential gains, more of which we’ll be highlighting in future posts.

Today we won’t focus on STIPA, but rather on STIPA’s Social Marketing Officer, Peter.

I was able to spend an afternoon with Peter when, because of a torrential storm and subsequent power outage, we were left unable to work on the desktop computers. I took it as an opportunity to learn more about the culture of M-PESA to share with those who have little, or no experience with it.

Peter is a young Kenyan who was recently living in Nairobi but has now relocated to Kisumu to work for STIPA. Peter’s been using M-PESA now for over five years and considers it a valuable tool, but doesn’t think about it that often, as it’s just a part of his daily (or monthly) life. When I spoke with Peter it was mid October and Peter currently had 12,000-15,000ksh (roughly $130-$165USD) on his M-PESA mWallet.

M-PESA: part of daily life

I was interested to find out how Peter uses M-PESA in a typical month and question why he’s not using it in other ways. Peter described a typical monthly interaction for me. He’s paid his monthly salary through a direct deposit into his bank account, and withdraws approximately 30,000ksh (roughly $330USD) and deposits that amount onto his M-PESA account by visiting a large M-PESA agent at a nearby shopping centre. Peter shows the agent his government issued ID, confirms his mobile number, signs the register and hands over the cash. He then receives an SMS from M-PESA confirming his cash was deposited onto his mobile account. The entire transaction takes approximately 1 minute.

Peter makes this deposit onto his M-PESA account so that he can:

1. Pay for bills (electricity, water, internet) by sending directly through M-PESA to his accounts using PayBill.
2. Withdraw from M-PESA at a later time to buy food and house supplies from the grocery store.

Interestingly, he does this because it’s easier, and cheaper for him to withdraw using M-PESA than it is to use his own bank card for numerous transactions, plus its impossible to pay electricity/water/internet bills using his bank card. Nonetheless, he does not pay for groceries at the supermarket using mobile money, even though it’s possible at most large grocery stores.

When asked why he doesn’t keep his entire monthly salary on the M-PESA mWallet, Peter had two very quick answers:

1. It’s too easy to spend money on M-PESA, if he kept it all on his M-PESA mWallet it may be spent very easily and quickly decrease his savings.
2. It’s impossible to build up credit through M-PESA. Peter said he’s looking at applying for a loan soon, for which he will need a credit history.

M-PESA and credit

At the moment, (though it’s changing) loans are only available if money is kept in an accredited bank account. This is one of the major incentives of joining a Sacco (Savings and Credit Cooperative) since you’re able to rely on other members of the group for loans which exceed your personal worth.

Access to credit—even microloans of just $10-$50 USD—are hard to come by for the unbanked. It’s one of the major issues plaguing the poor who aren’t able to invest beyond their personal available capital. This difficulty, a kind of financial glass ceiling, is being addressed in unique and innovative ways by various actors. Two such companies looking to change how last mile, low-income populations access credit using mobile are Inventure and First Access. Inventure is able to predict creditworthiness by helping low-income people with tracking their finances by tracking financial information on an Android application. When applying for a loan, a user can point to their recent savings and expense history as evidence of creditworthiness. After a quick multiple choice questionnaire, the loan can be approved instantly. First Access, on the other hand combines financial and recent mobile data usage, such as the amount of the person’s most recent top up, number of days with no mobile traffic, and other data as proxy indicators to predict creditworthiness through an algorithm.

Everybody’s worried about phantom fraud

When asked what could be improved about M-PESA, Peter quickly responded that issues to deal with fraud need to be addressed - although he personally has not ever experienced fraud while using the mobile money service. This immediate, almost unconscious answer unaccompanied by a heart-wrenching story, caught my interest. Is the real threat of fraud as high as it’s portrayed? How often does it actually occur?

After my conversation with Peter, I spoke with five other employees, none of whom had ever personally experienced fraud with M-PESA, but all of whom insisted it was a concern. A quick internet search revealed that, according to Bob Collymore, CEO of M-PESA’s parent company Safaricom, M-PESA fraud occurrence for customers is around 0.002 percent, and for merchants it is 0.007. For comparison, the average credit cards fraud rate is 0.05 percent (The Star). That means that out of every 1,000 people, only 2 will experience fraud over M-PESA as opposed to 50 with a credit card.

If I asked a different group of young people about their concerns with credit cards, would the first concern listed have been fraud? It seems unlikely—but it does seem that unfounded concerns are deeply associated with M-PESA, something that clearly needs to be addressed if we want to see stronger growth in take-up and usage.

But are we overestimating the ‘rural float’ challenge?

The last interesting piece of information Peter volunteered, was that in the last 3 months, he can’t remember ever having a problem withdrawing or depositing cash at an agent. Mobile float issues arise when the agent either does not have enough cash to hand out, in exchange for the M-PESA sent, or when someone is trying to deposit cash at an agent and the agent does not have enough on their mWallet to send to the depositor in exchange for the cash. The ‘rural float’ problem has often been cited as the reason mobile money fails to gain traction in rural areas. However, Kisumu is a large city and Peter tends to visit only large agents located at the supermarket, so it may just be an area where the mobile money ecosystem is functioning in equilibrium.

So what does this all mean?

To me, it means mobile money is not where it needs to be to support the transition to a more cashless society - even in Kenya, arguably the world’s most developed mobile economy. More financial products need to be available through mobile money for users to access loans, better savings mechanisms need to be in place to encourage saving and prevent irresponsible spending, and there needs to be more focus on user education around the true incidence of fraud and how to prevent it.

The biggest barrier to using alternatives to cash, whether in Kisumu, across Kenya or globally, is the need for behavior change—awareness of alternatives and the courage to test them. But this is also at the core of its potential—in Kenya, as in much of the developing world, youth represent the majority of the population, and as this tech-savvy market segment grow into their purchasing power and begin to try out new mobile money products, we predict an explosion of mobile financial products to serve them. Some of this will be possible through products targeting young people specifically. Others will offer benefits such as interest-bearing mobile savings accounts, the ability to develop a credit history, loyalty vouchers, and links with social media. Critically, they will need to be fast, efficient and easy to use.

Competition in the market will be beneficial—the more businesses there are thinking about how to attract new customers and grow traffic, the better. This will spur more and better targeted ideas, and allow for rapid prototyping of new and edgy product ideas. Interoperability is also at the core of this - the easier we can make it for people to innovate on top of service infrastructure like M-PESA, the faster uptake and throughput will grow. APIs will be good news for everyone.

We aren’t there yet, and have some significant hurdles to cross, but understanding how mobile money is being used by the average person gives us an insight into where effort needs to be focused in order to increase usage, desirability and trust.