How does technology help develop the most effective pre-delinquency strategies?

The common conception of collections is as a historical activity. A debt was owed and is now delinquent. So it’s the act of rectifying a historical account.

This conception is right, to an extent. That’s certainly a large part of the process. But it’s not the full picture. We’ve written before about how collections are one step within the customer lifecycle. Not just a way to recoup debts, but to retain customers.

But drill deeper, and you’ll find that debt has its own steps embedded within it: A more modern approach to collections focuses on pre-delinquency, too. That is, the treatment of accounts before they exceed limits or miss payments.

By recognizing the warning signs of delinquency you can proactively help your customers before the debt advances into delinquency. This is important for a few reasons. Most obviously: it spares you money and resources by quickly curing what could have become an impaired debt.

And now, thanks to IFRS 9, this takes on new importance: The new impairment model within IFRS 9 requires recognition of full lifetime losses more quickly. That means a greater emphasis on pre-delinquency to halt the advance of an account from ‘stage one’ (a 12 month expected credit loss) to ‘stage two’ (a lifetime expected credit loss).

So if early identification of credit-stressed customers is more vital than ever, how to do it? Jump too early or act too aggressively, and customers will feel harassed. Move too late and, well, you’re too late.

The importance of pre-delinquency strategies

No one customer is the same. Every person, every household will be at different steps in their debt lifecycle. A sudden unexpected event, the loss of a job - there are numerous reasons why a customer’s finances can suddenly worsen and veer towards delinquency.

You don’t always know the ins-and-outs, but it’s possible to accurately identify when an account begins to slip. And identifying someone who is pre-delinquent (not quite in the red, but in the orange) gives you control over how best to help them.

Collections software allows financial institutions to rate risk and flag the likelihood of delinquency and segment portfolios based on recovery needs and risk scores for more effective collector targeting.

Specifically, customer segmentation means identifying the right predictive metrics and moving swiftly when a customer account begins to exhibit these signs. Examples include changes in repayment amounts, increase in usage, late payments, or cancellation of the direct debit.

The beauty of accurately identifying pre-delinquency is it allows for less aggressive communication. Your collections department has a head start and your collections system allows you to clearly define the stages of communications and your tone. Far from being threatening, your tone can be educational: offering advice or giving options.

But while segmentation and data are a good step, without testing your assumptions, it’s impossible to know what approaches are working. Technology allows you to explore different scenarios and provides the tools for testing frameworks and measuring success.

Your approach to identifying and tackling pre-delinquency can be scientific, not just based on a hunch. EXUS’ champion/challenger facility is an example of what that might look like in practice. If you want to test a new workflow, a champion/challenger session lets you try it on a limited set of accounts.

Once you submit the workflow, you can edit it as you see fit. In the end, you can see relevant reports and analytics to give you a feel for the success of the strategy. If it leads to fewer delinquencies, then roll it out to more accounts.

The next step

The challenges of pre-delinquency also point to an exciting solution: Artificial Intelligence. While we don’t yet have an AI that can pass the test famously formulated by Alan Turing in 1950, we have seen remarkable advances in the field.

We’re far from the golden age of AI, but we are living in the golden age of AI-enhanced productivity. AI transformed Google Translate from a stilted laughing stock, into a service with an uncannily human appreciation for artful language and nuance.

Outside of Google’s advances, the immediate future of AI is as a colleague. Narrow domain AIs that can perform pre-defined tasks - like, perhaps, running tests on a batch of pre-delinquent customers. While still somewhat in its infancy, AI is nonetheless breaking new ground in the collections' ecosystem. And it will only improve the more data it collects.

While champion/challenger offers standard A/B testing of individual bits of functionality, machine learning and AI will be able to test hypothesis themselves, refining the collections process based on the best result. The time-consuming iterations of test and refine can easily be handled, human oversight being the only non-machine task.

Conclusion

Ultimately, the goal of all this, the one true metric of success, is to speed up your collection process.

You are now able to help your customers before their problems advance too far. Accurately diagnosing and treating the signs of pre-delinquency is the best step you can take. But how can spot the signs early on?

If you are still relying on antiquated systems, clumsy workarounds, or excel spreadsheets, you simply don’t have the insight you need to answer that question. The only way you can achieve this is not only through investing in tech -- but investing in the right tech.