Are credit bureaus still relevant for the underbanked in the age of alternative data?

Traditional credit bureaus have long been the backbone of formal lending decisions, tracking repayment histories from loans, credit cards, and utilities. They were were built for a specific kind of person: salaried, formally employed, already in a banking relationship. Their architecture reflects this: they aggregate loan repayments, credit card statements, utility bills, all events that presuppose prior access to formal financial services. Yet, for the roughly 1.4 billion adults who remain underbanked or unbanked (think informal traders or rural farmers in SubSaharan Africa, Eastern Europe, Southeast Asia, Latin America, etc.), they create a a well-documented trap: you need a credit history to access credit, but you can only build a credit history by accessing credit.

Perhaps unwillingly but still actively, the system perpetuates exclusion by treating the absence of formal data as evidence of risk, rather than as a symptom of invisibility. Meanwhile, these same populations generate high-signal behavioral data every day through transactions, payments, and financial decisions that traditional bureaus are architecturally blind to.

Enter mobile money data and ecosystems across Africa and Asia, flipping the script with rich, real-time behavioral insights. These platforms provide a qualitatively different information layer in near real-time: transaction velocity, airtime top-up consistency, peer-to-peer cash-flow patterns, merchant payment frequency, etc. that serve as a high-signal proxy for financial character and liquidity.

The data generated is not a proxy for creditworthiness; in many cases, it is creditworthiness.

Enter mobile money data and ecosystems across Africa and Asia, flipping the script with rich, real-time behavioral insights. These platforms provide a qualitatively different information layer in near real-time: transaction velocity, airtime top-up consistency, peer-to-peer cash-flow patterns, merchant payment frequency, etc. that serve as a high-signal proxy for financial character and liquidity.

For hundreds of millions of people, M-Pesa, bKash, MoMo or Wave is where economic life actually happens. The data generated is not a proxy for creditworthiness; in many cases, it is creditworthiness: more honest, more current, dynamic and more granular than anything a traditional bureau holds. It reveal how individuals manage money and build resilience to shocks in their daily lives. This shift from debt-based scoring to behavioral identity allows for a more accurate and inclusive assessment of risk at scale.

Lenders who have piloted alternative scoring models in these markets have found that behavioral signals from mobile money significantly outperform thin-file bureau scores in predicting repayment, particularly at the lower end of the income distribution where conventional models fail most.

In many contexts, alternative data is clearly better. Therefore, are credit bureaus irrelevant or suddenly obsolescent in a world where richer behavioral data already exists at scale? They need not disappear. They could evolve from passive repositories of formal credit events into active data exchanges that ingest, validate, and standardize alternative signals alongside traditional ones. Several bureaus are already moving in this direction, partnering with mobile operators and fintech platforms to broaden their data layer.

But integration raises its own complications: appropriate consent and privacy frameworks, as well as adapted governance models, should be set at the regulatory level, to avoid the cure being worse than the disease.