FinTech Seeks Improved Data Ecosystem For Credit Score Metrics

The managing director of Sofri, Mr. Paul Adebayo has stated that improved data ecosystem will harness credit score metrics in Nigeria’s credit industry.

He reinforced the need for robust data ecosystem to harness credit metrics even as he called on government to bridge data gaps by removing the hurdles and bottlenecks that hinders credit access in the country.

Adebayo who stated this during Sofri’s relaunch media parley in Lagos, noted that, without a robust data ecosystem, the financial sector will struggle to scale inclusive lending products.

According to him, the credit industry in Nigeria suffers from limited and low-quality data, which hampers the ability of lenders to assess risk and design effective credit products.

He, however, noted that, the recent moves by the government to promote the deployment of Global Standing Instruction (GSI) will encourage financial institutions to push out loans and deepen financial inclusion.

He explained that, GSI policy framework by the Central Bank of Nigeria(CBN) enables lenders to recover loans from any account linked to a borrower’s BVN.

“With GSI, if I give you a loan and you move your salary to another bank, I can still retrieve repayment as long as it is tied to your BVN. That is a game-changer for credit recovery and risk reduction,” Adebayo said.

Commenting on AI and social credit, Adebayo also revealed that, Sofri is integrating artificial intelligence and machine learning into its lending decisions to better assess borrowers, especially in the nano and payroll lending segments.

“We are big on public sector lending, and if you work in a private company or run a business, you can get a private loan. But we’re also expanding nano loans to meet urgent needs of N2,000, N5,000, or N10,000,” he said.

To assess such borrowers, he said, Sofri considers a broad range of alternative data points beyond just salary or employment history. ‘You and someone else might earn the same and work at the same place, but your credit behaviours are different,’ he explained.

Using machine learning models, Sofri now evaluates variables such as phone usage, location, education, and lifestyle patterns. These inputs help the company assign credit scores even in the absence of a formal banking history.

He disclosed that Sofri is exploring the use of social media metrics as part of its credit scoring model. ‘If you have 10,000 followers on Instagram and I have 2,000, the probability that you will default is lower, because your social capital is higher,’ he said.