Jay CabozBy Jay Caboz|January 18, 2022|13 Minutes|In Billionaire Tomorrow

Billionaire Tomorrow

" Who needs a bank when you’ve got a phone loan?"

Samuel Njuguna saw a gap in the market where even the big banks feared to tread - loaning money to people who don’t even have a bank account. Throw in a dash of A.I. and you have an entrepreneur who can audit you through your cell phone data.

It all started with an idea to help Kenyans without bank accounts borrow money. A problem shared by a staggering 43% of Africa’s 1.3 billion people who don’t have one either. Which leads without a credit score and leaves then shut out of the financial world.

For entrepreneur Samuel Njuguna this was a challenge. His team of co-founders in the fintech space, felt it was a problem they solve using technology to open up a flood of credit to those who need it most.

“The unmet financial need in Kenya is worth $19 billion while it is worth $331 billion in Sub-Saharan Africa. Only 15% of borrowers in Africa borrowed from formal institutions,” he says.

“We have had personal experience where our parents who do not have formal bank accounts would have an issue getting loans and in turn, they would end up borrowing from friends or loan sharks.”

The lightbulb went on when Njuguna came across some A.I. called machine learning. It creates credit scoring, even if you don’t have a bank account, by drawing data from your cell phone.

This was the birth of Weza in 2016. It opened a vast credit pool to Kenyans, and is now spreading to countries Uganda, Ethiopia, Ghana and Cameroon- all markets ignored by established credit scoring systems, says Njuguna.

“Weza offers credit scoring and fraud detection services to all sorts of businesses. We are making use of alternative data to inform the credit scoring.”

One of their farthest clients reached out to them in the strangest of ways, all the way from Haiti.

“For the Haiti engagement we got covered by an online magazine and someone from Finca Haiti reached out via LinkedIn for our services.”

Like fintech companies Flutterwave and Naked Insurance that are disrupting the financial and insurance sector, Weza bases its business model on harnessing data from smartphones.

The reason?  More people in Africa have a cell phone than a bank account. Africa is the world’s fastest growing mobile phone market. Experts forecast that by 2025, there will be 615 million people, equivalent to 50% of the region’s population, a report by GSMA found.

A 2018 report by McKenzie, found banks are now harnessing Africa’s widespread mobile-phone coverage to create low-price offerings and innovative distribution models.

“As of 2017, there are almost 300 million banked Africans, up from 170 million in 2012. By 2022, we project 450 million banked Africans. By 2022, close to half of Africans will be banked, compared to just over one-third today,” the report found.

This is despite Africa still having many challenges from low-income levels in many countries; widespread use of cash in most economies; poor coverage of credit bureaus; to fraud.

Which is where Njuguna and Weza came in.

“Examples of this data being mobile phone data such as airtime leaded, mobile money transactions and payment data,” he says.

Growing up, there was more resistance than I expected because the norm is to finish school and be employed

Samuel Njuguna

When Weza started in 2016, Africa’s banking sector was a completely different space. Around 57% of people didn’t own a bank account.

The company took its first steps loaning their own money as they tried to understand the complex world of mobile money.

“For us to understand the credit scoring model, anti-fraud models and the end-to-end infrastructure we started off by lending our own money…However, we got a number of companies asking whether we would implement the same platform for them. The key reason the companies were asking for the solution was we had come up with credit scoring models and anti-fraud models which were robust. With the requests and our capability, we realized we can offer our platform under a business-to-business offering,” says Njuguna.

They soon found out there was a massive demand for their credit scoring system, which used A.I. to detect fraud – one of the key reasons why many unbanked were denied access to loans in the first place.

“Fraud in the space is rampant and the form we have is one that manifests as identity theft in digital lending. A fraudster registers as person B, registers on the platform of the lender as person B and once they get the money, they throw away the  phone number they borrowed with alternatively they commit fraud on other platforms. Lenders will think that the person B committed a default but in actual sense it was a fraud case,” says Njuguna.

The company pivoted to offering their machine learning algorithms that uses the data available on customers’ cell phones to predict creditworthiness to catch out and identify theft.

“Identity theft is a huge problem in digital lending. In our initial roll out we knew it was present and we had seen other lenders hard hit by it. Our approach was to identify how it occurs and mitigate it. In the process we incurred a loss of revenue through the fraud, but we learnt how to mitigate it by checking the data sitting on the user’s device. The way we can detect a fraudster is by checking whether the SIM has been getting constant communication coming in and going out, a fraudster phone normally has messages coming in but very few going out since he doesn’t communicate to friends with that line. We use this and other key data considerations to determine authenticity beyond SIM card registration details,” he says.

Njuguna’s introduction to the world of technology came a lot later than most.

“My background is in Computer Science. However, I grew up in the village and my first interaction with a computer was in high school. I loved the interaction with the computers and it so happened we had a very supportive teacher and this led to interest in the space. Going to university I knew that’s what I wanted to pursue,” he says.

As for other young Africans looking to start their own businesses, Njuguna advises you to dive in and learn from experience.

“Growing, there was more resistance than I expected because the norm is to finish school and be employed. It’s only now that my parents are seeing working on a start-up as an alternative route and this is because this channel is growing in popularity in Kenya,” he says.

“You can try and read all books about entrepreneurship or listen to a million YouTube videos but nothing beats experience. Jump right in and earn your stripes.”

Much like how he learned to adapt to the world of computers, so too has the business which has its home base in Nairobi.

“We were doing side gigs to get the money to provide as lending float for the service, and we eventually got some seed capital. This allowed us to understand the risk involved and their mitigations before we could offer the solution to other businesses,” says Njuguna.

Weza now offers flexible offerings for both B2B and B2C clients.

“If the client already has clients and wants to offer lending as a value-add service, we utilize the data the client already has to formulate credit scores and if it’s not enough we do advise on what other data to collect. We then layer the lending infrastructure from the dashboards to the end user apps on the clients services,” says Njuguna.

“For a client who wants to start lending either to the public or employees of other companies and has not been offering any service, we also customize our infrastructure to collect data for credit scoring and offer interfaces for the end user to interact with. In other cases, we just build credit scoring infrastructure for the lender.”

Weza is currently running another seed round looking for $1,5 million. In the last months, it has secured a new partnership to provide services in Ethiopia and is looking for similar partnerships in other countries as a route for organic market expansion.

There you have a business born from a gap in the market because banks weren’t brave enough to lend to people who had a cell phone and little else.

Njuguna says there are three ingredients to building a successful lending platform:

1) An astute credit scoring model – to make sure you credit score the clients in an optimal way based on the data type, data size and consideration of other data sources to augment what you have. This is meant to reduce defaults and late loan repayment.

2) Strong anti-fraud is to ensure you don’t have cases of fraudsters siphoning money off the platform leading to loss of money.

3) User interface design is crucial in developing markets since the people who are using your platform have different literacy levels and the platform should communicate to each and every one of them.