Back in the 16th century, retailers offered different sales channels and products to customers of different economic classes. Evidence of segmentation can even be traced to the Bronze Age, when merchants segmented their markets based on geography. Today, this age-old concept is more important than ever, especially for those working in credit management.

The assumptions credit risk professionals made up until now might not be relevant anymore, and the way they look at their customers needs to change accordingly.

Key steps to an effective segmentation

Here are five new ways to look at your customer base, and prospects, to ensure you’re properly managing your risk and, at the same time, not leaving money on the table.

  1. Run mini-models in parallel to help you achieve optimal results. The Covid-19 crisis will impact customers’ financial stability in various ways. Thus, a more granular segmentation will yield more accurate results. Ensure your risk modeling platform allows you to run these in sync so you make the most out of this micro segmentation approach.
  2. Take into account traditionally-used internal data sources like salary, recurring expenses and free cash flow, as well as alternative ones like channels data, devices used to communicate with your organization and other pieces of data that might be relevant for your segmentation.
  3. Consider new data variables that demonstrate an individual’s or an industry’s risk profile. An example can be eligibility for a government stipend, which some governments are now offering.
  4. Understand the growing impact of certain characteristics: Maybe now is the time to emphasize specific financial behaviors or occupations?
  5. Embrace machine learning to uncover hidden yet meaningful segmentation factors. Sometimes, the not-so-obvious characteristics are the ones that have a major impact on your results.

Act now, and act fast

You’ve heard it in many contexts in the past few weeks, and you’ll probably keep hearing this again: the world is changing. Make sure the way you approach credit risk management adapts to the new reality. Use more advanced modeling techniques, especially those leveraging machine learning, even if only to guide your sanity checks. The cost of mistakes has never been more painful.