In our previous blog post we looked at data analytics in collections and the expected change in performance.
Strategically, data analytics drives operational execution, but the question remains: where do we start? In this blog post, I outline the 3 steps to building your own data-driven collections strategy.
There is no “one-solution-fits-all” approach; a strategy that works for one company, does not necessarily work for another. So, let’s first agree that each approach or strategy needs to be tested, and when testing a strategy, a testing methodology is required.
The champion/challenger method is a contest between “the champion” (the current strategy with the current analytically derived account segmentation, prioritisation and treatment framework) and “the challenger” (the proposed strategy with different analytically derived account segmentation, prioritisation and treatment framework) where the winning strategy produces the best results and therefore becomes the new champion during the next test run.
So let’s look at the steps to build our own analytically derived strategy.
Step 1: Decide what you want to analyse
Begin by deciding what you want to analyse. For example, do you want to analyse Risk, such as propensity-to-pay/propensity-to-roll?
Then ask “Why?” Do you want to support operational efficiencies while increasing collections and recoveries?
Finally, ask yourself, “What do we need?” Do you need a Risk Segmentation of your delinquent account population?
Collections Risk Segmentation allows for risk-tilted treatment and intensity, and it enables prioritisation of in-house effort vs. handover to an agency.
Typically risk segmentation is based on either risk scoring or derived by means of a simple segmentation tree, and it generally concludes in 3 to 4 risk bands, e.g. low, medium, high and very high. Also, risk segmentation will often be calculated on an account when it first enters collections and will then be updated on a monthly basis.
Step 2: Decide on the attributes that may influence the risk possibility
Next, decide on the attributes that may influence the risk possibility. To ensure adequate risk ranking, risk segmentation should incorporate extensive data sources:
- Account level data, e.g. balance, past account performance, etc.
- Customer level data, e.g. time since salary payment, delinquency history, total balances, debt to income ratio, Government Sector, Private Sector, other deposits, etc.
- Credit bureau data (bureau data can provide significant additional lift), e.g. payment performance at other institutions, level of recent enquiries, total debt commitments, etc.
Step 3: Build a scoring model or simple risk segmentation tree with the identified attributes
Now that we know our debtors’ “risk”, we are now ready to build our operational execution strategy.
Should you require any guidance or data analytical support in the development and execution of your data analytics driven collections strategy, please feel free to contact us to discuss your requirements and answer any questions you may have. You can also learn more about our Collections & Recoveries Solutions here.
About the author
Perry de Jager: Head of Collections and Recoveries Solutions
Perry has been with Principa Decisions since 2015 dealing with Collections and Recoveries solutions for clients within South Africa, Africa and the Middle East.