Insights · Analysis · for financiers
Responsible credit scoring and the future of financial inclusion in Africa
New data can widen access to credit, or entrench old exclusions. What responsible scoring on trading data has to get right.
Mtaji Insights
First published 9 March 2026; edited for this site 28 October 2026. 2 min read.
Written by the Mtaji Insights desk and checked against Mtaji’s product terms and publication rules before release. Editorial standards
In brief
- For business credit, the most useful new data is business data: how a firm trades with its anchor.
- Responsible scoring uses commercial performance, explains what it measures and never gives a bare refusal.
- The credit decision stays with the financier, who answers to its regulator for it.
New ways of assessing credit are spreading worldwide. Nowhere do they matter more than in Africa, where the gap between economic activity and formal credit records is widest.
Why traditional models miss most businesses
Credit bureaus in developed markets work because formal financial activity is the norm: years of loans, cards and utility payments on file. Here, much of the economy runs outside those records. Our bureaus have made real progress, but they still see employees and established firms far better than the small businesses that create most jobs.
This is not a shortage of activity. The transactions happen every day. It is a shortage of systems that turn that activity into something a lender can rely on.
Alternative data, and its limits
Mobile money, utility payments and online sales are all being used to assess credit. Mobile money in particular shows how often money arrives and how steadily. But it says little about whether a business delivers on time, keeps its quality or can handle a larger contract. For business credit, we need business data.
Trading data: the missing layer
When an anchor runs its orders, deliveries, invoices and settlements through a platform, every step leaves a structured record of how its suppliers and distributors perform. Over time, that record shows reliability in a way no other source can.
The M-Score reads that record. It tells a business what it measures and what to do next, and it keeps no secret about which kinds of information it uses. It does not publish how they are weighed.
Responsibility comes first
Any scoring method can widen access or entrench old biases. A model trained on biased history can repeat it, and a decision nobody can explain cannot be challenged. These are documented harms, not theoretical ones.
So the rules we hold ourselves to are plain. Use commercial performance, not social media, phone usage or contact lists. Tell a business what is measured and how to improve. Never give a bare refusal: give the gaps and a person to talk to. Keep each person’s consent separate. And keep the credit decision with the financier, who answers to its regulator for it.
Explainability is not only ethical. It is commercial. A bank will not lend on a method it cannot justify to its supervisor.
Looking forward
Better data, honest methods and sound governance can together make the invisible visible. No method closes the credit gap alone. A trading record that a lender can rely on, built on consent, is a large part of the answer.