The next generation of collections modelling

As the Bank of England predicts that inflation won’t drop to two per cent until the third quarter of 2024, it is unlikely that the cost-of-living crisis will dissipate anytime soon.

This is particularly concerning for collections teams who play a heightened role during uncertain economic times, as they did during the Covid-19 pandemic, in ensuring their firms are supporting struggling customers as much as possible.

It is therefore essential for these teams to prioritise efficiency and effectiveness within their strategies and leverage analytics and modelling to help improve collections performance.

Existing approaches

The use of analytics and modelling in collections strategies has been around for many years. These approaches include:

  • Collections segmentation: enabling the tailoring of treatments based on risk profile and likelihood to engage
  • Propensity-to-pay modelling: quantifying an individual’s repayment likelihood to identify which treatment path is most appropriate
  • Champion challenger frameworks: determining the effectiveness of one strategy or form of contact over another.

These approaches are key to ensuring the effectiveness of a collections program. However, when assessing the success of a programme, effectiveness is not enough. Efficiency must also be evaluated. And all this within the context of Consumer Duty expectations.

Effectiveness relates to how an activity or strategy may influence the likelihood of an action, whereas efficiency asks, “was it worth it?”. For example, phoning or mailing a customer may result in them making a minimum payment, but does it offset the costs of operating the call centre and sending mail? An efficient strategy should be designed to achieve its desired objectives while minimising costs to ensure good customer outcomes.

Leveraging attribution modelling

Effectiveness and efficiency can both be achieved by understanding which treatments are producing the most successful outcomes. Attribution modelling is usually deployed within marketing strategy, it can be used within collections to measure the incremental impact of a particular treatment or combination of treatments.

From a mathematical perspective, both collections and marketing strategists are seeking to solve the same problem. In marketing, businesses want to understand how a combination of touchpoints can persuade potential customers to buy. In collections, lenders are interested in how their outbound treatments can encourage existing customers to engage and make payment.

Attribution techniques can take several forms, for instance:

  • When combined with customer segmentation, lenders can see how different treatments at different times could provide a more efficient collections strategy.
  • By testing various outbound channels, containing different messaging, including styles and tones of voice, lenders can identify which messages will be most successful with which customers.
  • Finally, lenders can establish which contact channels are not cost effective for specific customer populations. the spend can then be repurposed to activities and strategies that provide the highest return.

Harnessing the insights

Existing approaches, such as customer segmentation and propensity-to-pay modelling, can be effective at increasing recoveries and cure rates, but don’t provide a complete view of what’s most efficient.

Attribution techniques can be used to measure the incremental impact and return of a particular treatment or combination of treatments. Armed with this knowledge, lenders can boost the efficiency and effectiveness of their collections strategies, all while improving operations and having a positive financial impact on their business and supporting customers in difficulty.

For more information, visit or contact Matt Triggs at