The collections industry is running on a dignity deficit: Ananth Shroff

The rush in retail lending and the associated high stress in loan quality has put the spotlight on debt collections. With the Reserve Bank of India insisting on responsible recovery and customer dignity, an empathetic collections model is taking shape. Ananth Shroff, cofounder and chief executive officer of DPDzero, which uses artificial intelligence (AI) for debt collections, interacted with Raghu Mohan via email. Edited excerpts:

 

Why do you think the collections side has to be reimagined?

 

Retail credit has exploded, tech has scaled while disbursals have got slick, and digital lending has seen a surge of approximately 132 per cent (compounded annually) in the last few years. However, collections have stayed where they were — outsourced to agencies and fragmented in approach. The outdated approach of increasing outreach frequency or intensity of communication only reduces the creditor’s ability to contact debtors to resolve delinquencies. It is important to understand the psyche of today’s borrower — they are digital natives, tech savvy, and sensitive to the kind of experience they receive from the lenders when it comes to loan repayments. The answer to resolving collections is not to keep waiting to write off bad debts in due course of time every financial year but to have a change in mindset shift in how collections is perceived, discussed and tackled in the boardrooms. Your future valuation depends on today’s book quality. Your brand reputation gets judged not when you disburse but when you collect.

 

What of the boots-on-the-ground approach to collections: Is it beyond the sell-by date?

 

The key challenge we see is that they chase payouts and do not bother to understand the borrower. The lender is busy building relationships with the vendors, instead of borrowers, which to me is exactly opposite the way it should be. The on-ground approach puts emphasis on process-first and borrower later. And that is what we want to change, because the collection industry is already running on a dignity deficit, and this manpower-heavy approach only adds to the trouble. Loan repayments cannot be forced; it needs to be inspired. And that can happen only with a relationship-based approach with the borrower.

 

The entire chain from onboarding to collections works in silos across lenders. How is this issue to be tackled?

 

Interestingly, this problem didn’t exist back in time. Talking about the early days of micro-finance institutions in 2007-09, while they were financing the underserved, there used to be only one person responsible for disbursals and collections. The repayment rates were great at 98 per cent and loss rates were less than 0.5 per cent. Today, that borrower journey and intelligence is decoupled and fragmented, while the sales team is largely concerned about disbursements, and the collections team is thinking of ways to collect more. This problem has only aggravated with the rise of digital lending in the past couple of years, where we lend without having a 360-degree view into the borrower profile or their financial behaviour. A lot of lenders today are contemplating how to not move away from that intimate model so that the human connection is not lost. Today, debt collectors sit on primarily four data points: Name, contact, number, due date and loan amount. There is a lot more data available with the institution to have a complete borrower picture.

 

Can AI fix collections?

 

We believe that each borrower is unique and so should be the collection journey. AI can create personalised, context-aware messaging templates that make borrowers feel spoken to rather than broadcast at.

 

What of the costs involved in digital collections? As in, how is this aspect to be calculated and shared across the parties involved — be it lenders, sourcing agents and collection agencies?

 

The cost (of collections) depends on the quality of the borrower and productivity of the collection teams. A borrower with a Cibil score of 780 is different from a borrower with a score of 550 in terms of the repayment behaviour exhibited in the past. And a borrower with a fixed obligation to income ratio (FOIR) of 60 per cent is more likely to pay than a borrower with FOIR of 85 per cent. These data points help us segment the borrowers into various risk cohorts and custom tweak our collections strategy accordingly.