Cracking Informal Credit: How Data Is Rewriting Risk in Nigeria’s Untapped Economy

By: Amina Yusuf

For millions of Nigerians working outside the formal economy—market traders, bus drivers, tailors, and food vendors—access to credit has long felt out of reach. Lenders, constrained by rigid scoring systems and paper-thin records, often see these individuals as unbankable. But that perception is beginning to change, thanks to new advances in data science that are reshaping how risk is assessed in Nigeria’s financial system.

Between early 2018 and mid-2019, a consulting-led initiative involving several financial institutions launched a pilot model that used behavioral data—such as mobile phone usage patterns, airtime top-ups, and digital wallet activity—to evaluate creditworthiness. At the heart of the initiative was data scientist Samson Edozie, who played a lead role in designing the alternative scoring framework.

Rather than rely on payslips or formal credit histories, Edozie and his team trained algorithms on patterns derived from thousands of anonymized transaction profiles collected across urban and peri-urban centers. These models looked for proxies of reliability: consistency in cash flow, mobile recharge behavior, and even bill payment frequency.

“The traditional credit score was built for a different economy,” Edozie told analysts at a private sector roundtable in May 2019. “In Nigeria, where so many operate informally, we needed a model that speaks to behavioral trust, not just financial records.”

The pilot was rolled out with a cluster of microfinance partners in Lagos and Aba, reaching just over 18,000 users within the first five months. Early results were promising: approval rates nearly doubled, and default rates dropped by nearly 40% compared to traditional onboarding processes. Informal workers who had never received a formal loan were now gaining access to working capital in under 72 hours.

One Lagos-based lending partner, requesting anonymity, described the model as “a turning point” in how they think about risk. “We used to reject applicants simply because we couldn’t find them on any formal database. Now we understand their financial lives in a more relevant way.”

This data-driven approach is not only increasing access—it’s influencing how financial institutions structure their products. Several lenders are now tailoring loan tenures, repayment frequencies, and interest rates based on user-level behavioral clusters developed during the pilot.

While broader adoption is still in progress, Edozie’s contribution stands out for one key reason: it brings nuance to risk scoring in a country where nuance is too often ignored.

As Nigeria pushes toward its national financial inclusion goals, this type of work may quietly become one of the most powerful drivers of change—by building systems that see people not as risks, but as participants.