Digital debt collection makes strides as lenders use tech to cut visits

“We currently manage collections for over 98 million retail and small-business loan accounts, covering a loan book of about $250 billion,” says Rishabh Goel, cofounder and chief executive officer (CEO) of Credgenics, a software-as-a-service platform for debt collection. With more than 700 million smartphone users and over 450 million UPI users, digital access is everywhere. “WhatsApp, used by over 550 million of us, is now a default channel for collections and legal followups,” points out Nitin Purswani, cofounder and CEO of Medius Technologies. And that “over 25 million people paid EMIs digitally in May 2025. Many of these are missed EMI recoveries”. 

 

For a business that’s taken huge strides in recent years, mirroring the growth in retail credit, there’s hardly anything by way of a sizing study on it. But what’s dawned is that digital collections help cut the huge cost lenders incur when they send recovery agents to customers. A single field visit to a customer can cost anywhere between ₹250 to ₹500; and there’s no surety that this will lead to a settlement even after a few tries. Now multiply this with millions of retail customers under stress who have to be called on — it adds up to a bomb. Worse still, the incentives offered by lenders to the agencies engaged in recoveries are directly linked to their strike rate — how many recoveries they close out. This leads to other headaches: it puts lenders at a good chance of violating the Reserve Bank of India’s norms to prevent customer harassment; and by extension, the legal consequences and reputational risks given the huge investments in financial brands. 

 

According to the Financial Stability Report June 2025, at an aggregate level, the per capita debt of individual borrowers has grown to ₹4.8 lakh in FY25 from ₹3.9 lakh in FY23. This rise in per capita debt has been mainly led by the higher-rated borrowers. Among broad categories of household debt, non-housing retail loans, which are mostly used for consumption purposes, formed 54.9 per cent of total household debt in FY25 and 25.7 per cent of disposable income as of FY24. Moreover, the share of these loans has been growing consistently over the years, and their growth has outpaced that of both housing loans and agriculture and business loans.

 

Another matter of detail. Even as unsecured retail lending has moderated — it forms 25.0 per cent of retail loans and 8.3 per cent of gross advances — its asset quality has relatively weakened compared to the overall retail portfolio — gross non-performing asset ratio at 1.8 per cent vis-à-vis 1.2 per cent in March 2025 — especially in respect of private banks. On the other hand, the SMA (special mention accounts) ratio — an indicator of possible stress build-up in loan books — has risen, led by state-run banks. 

 

The digital debt collection mode is gaining traction. “There is a clear inflexion point where digital collections are extending into small towns and even villages, driven by the digitisation wave in financial services,” says Yamini Bhat, cofounder and CEO of Vymo, a technology solutions provider in the trade. She points to RUGR, a startup, and CSC-VLE, a government programme for village-level entrepreneurs, rolling out vernacular digital payments and collections in rural areas.

 

Is the approach customisable? “It is, but not to the extent required,” says Seema Prem, cofounder and CEO, FIA Global. She highlights the data blind spots: Many rural and low-income borrowers are still “thin file” or “new to credit”. “Without adequate cash flow, behavioural, or repayment data, customisation risks can turn into mispricing or over-simplified underwriting. Credit models built on urban salaried behaviours often misfire in Bharat segments.”

 

She makes a case for Unified Borrower Exposure Registry: a regulated, real-time registry that consolidates borrower-level credit exposure across banks, non-banking financial companies, and digital lenders — including those offering buy-now-pay-later or short-tenure loans. This would help lenders calibrate offers to actual repayment capacity, reducing the risk of invisible debt build-up, and for institutionalising debt resolution protocols for low-income borrowers.