Harnessing AI for tailor-made debt collections to improve customer experiences and reduce NPAs

The Indian economy is on a growth path and moving beyond the pandemic induced disruptions. Retail lending, as a key enabler, continues to increase over the past few months and has surpassed the pre-pandemic numbers. Collections, which forms a crucial part of the lending lifecycle, is an area that needs to be ready to support this profitable growth journey. The renewed focus on making debt collections considerate for borrowers has highlighted the need for an overhaul. The recent regulatory guidelines on digital lending and FinTechs highlight the cautious approach to manage the future delinquency risk and protect borrower’s interests. Collectively, these challenges are fuelling the need to automate and revolutionize the industry’s debt collections approach for the future.

In most cases, debt collection strategies remain complex, inefficient, and outdated. The inadequacy of the traditional methodology to anticipate which debts would be collected and which strategy is most effective, puts lenders behind the evolution curve. Using a one-size-fits-all remedy without changing tactics based on outcomes has led to lower rate of recoveries and huge cost to collect. With the emphasis increasingly on enhancing the customer experience, cutting operational costs and scaling up, banks and other financial institutions are looking to improve the efficiencies and effectiveness of their debt collections.

Using ML and AI powered technology platforms, banks and other non-banking finance companies are now recreating the customer experiences in collections and reimagining their operations that support customers and drive better and faster resolutions. Deeper insights into delinquency risk and how to handle at-risk accounts are being used to adopt a tailored approach.

A modern customer-centric approach to collections

Traditionally, empathy and debt collections have not worked in tandem for a variety of reasons. But this is changing very fast as lenders realize the growing need to transform their collections to support growth.  By using data analytics and AI, and designing algorithms with predictive models in innovative ways, collections leaders are revisiting their end-to-end customer engagement strategies.

AI-powered collections technology platforms bring together the power of data intelligence, automation, and digitization to provide a long-term sustainable solution. They not only enable lenders to adopt an assistive, nudge-based approach but also help contextualize and personalize the engagements. Sophisticated algorithms can assist banks and other non-banking finance companies in determining the customer’s preferences, challenges and, behavioural patterns to align the collections strategies accordingly. This is a huge leap from the traditional approach, which is mostly focused on lender’s urgency to collect without giving due importance to customer experience.

Holistic digital enablement of operations

Digital communication channels are highly matured today, armed with comprehensive intelligence in terms of reach, integrability, customizability, and flexibility to suit the needs of all the stakeholders. For lenders, a holistic communications strategy that encompasses multiple channels including WhatsApp, SMS, Voicebots, Chatbots, IVRs, and Emails is a great place to begin.

ML models provide deeper client segmentation, recommend the best-suited communications strategy, and provide the necessary intelligence to tweak actions in real-time. The communications are personalized with relevant borrower information and vernacular language adaptation for improved response. Lenders are provided the last-level insights to fine-tune and highly optimize their collections approach. A comprehensive digital-first collections model has proven to be extremely efficient, cost-effective and agile for lenders.

Leveraging AI for an assistive approach

Smart conversational AI-powered virtual assistants improve customer experiences, reduce the time to collect, boost collections rate, and reduce the cost that goes into dealing with delinquent clients. These conversational AI bots, once ready, engage borrowers on a regular basis, assist with payments, and automate reminder notifications without the need for a human agent calling unless it is absolutely essential.

Voicebot calls that are pre-programmed and intelligent are superior because they let borrowers understand the situation, choose when and how they repay. Based on the borrower’s response to the Voicebot call, the next communication is automatically triggered with digital payment links, while the bot remains on line to support in successfully completing the transaction. This leads to a more humane approach to customer engagement versus the traditional mode of hostile calls and harassment.

Smart insights based strategy for collections

Artificial intelligence has the capacity to help make smarter choices on when to contact customers and how, based on their behaviors. It is now possible to design algorithms with predictive models using the information attained through data and machine learning. Debt collection strategies can be prioritized and applied built on the data derived from risk segmentation, optimal channel prediction based on past communication behaviour, and intent-to-pay. The data-backed decisions transform the overall collections approach with dynamic segmentation, real-time insights, and personalized communications.

Predictive analytics provides lenders with early warning signals on likely defaults and assists a change in the strategy based on the data discoveries. Based on real-time data, lenders can also drive appropriate actions across the lifecycle including pre-due stage, filed collections, litigation strategies, and track the performance of teams involved.

The time to reimagine collections

The increasing use of AI and ML in lending is opening up new focus areas for banks and non-banking finance companies. In debt collections, they are enabling capabilities such as early prediction of likely delinquencies, enhanced techniques of categorizing borrowers, and contextualized customer engagements to reduce defaults, speed up recoveries, and transform customer experiences. An augmented AI-powered approach when applied to debt collections, maximizes the potential that technology can have on the whole debt lifecycle, recreating personalized experiences for customers, driving faster resolutions, and reducing NPAs.

(The author is Mr. Anand Agrawal, Co-founder and CTO, Credgenics and the views expressed in this article are his own)