Skip to main content
Collections 101: Why You’re Still Not Using Analytics (Part 2)

In my last post, I pointed out the top reason I hear from collection teams on why they’re not using analytics — they think it costs too much. Actually, the cost is not that great, and the ROI is terrific.

Which brings us to Reason #2: Trying new stuff is hard.

Being in collections is hard enough work as it is. When people invest in a new collections system like FICO Debt Manager, it takes time and effort to get it up and running. Once it is, people get sucked into the day-to-day business of collecting debt.

Many people don’t understand the system they’re using well enough to know how analytics could be incorporated. And many don’t know what they’d do with analytics if they WERE in the system.

Let’s take these one at a time. If you have a collections or recovery system, it should enable you to incorporate scores into your strategies. Whoever sold you the system will be happy to show you how to integrate analytics. (If they can’t, you really need to talk to FICO.)

Now, let’s talk about the kind of scores you’d use. The two most commonly used scores in collections are these:

Risk scores rank-order individuals by their likelihood to make payments on time. FICO Scores are an example of these, based just on credit bureau data, but lenders tend to use risk scores based on their own customer data as well. This means that risk scores are often already available on accounts that go into collections, and people with higher risk scores have lower credit risk, and tend to be better payers. However, once borrowers are seriously delinquent, their scores tend to be low anyway and may not be the best way of prioritizing accounts.

Collection scores rank-order delinquent consumers by the likely amount they will pay. This is sometimes known as “expected collection amount.” These scores come from predictive models built specifically for collections, and these models look at things such as promise and payment history, payment timing, contact channel preference, collector notes and other internal and external data. Using these scores, you can collect and recover more by prioritizing the people who will give you the most bang for your buck.

The easiest way to get started with these scores is to set a “cutoff score,” and spend more time trying to collect from people above the score than those below it. Once you start to understand the value, you can use scores to make all kinds of decisions:

  • Given the number of calls that can be made in a day, how should calls be allocated to accounts?
  • Which accounts should be assigned to the best collectors?
  • How much money should be spent locating this customer?
  • Should this inbound call be routed to the IVR or to a collector?
  • What settlement rate should be offered to this account?
  • How much should be spent on letters?
  • Which accounts should be placed with a collection agency, and when?
  • How much should I pay my agency to collect pre-charge-off accounts?
You can start to do some really spiffy segmentation. Here’s an example:

Chart showing a strategy with and without collection scores

Still not convinced? OK, this is a tough crowd. Try this: Get FICO to score one of your worked debt portfolios, and I bet you a lot of money that you’ll find that the people with higher scores are paying you more money. But you had to work all the accounts to find that out – with scores, you’ll know this before you make any calls.

This is Collections 101: It pays to know the score!

For more about analytics in collections, download our white paper, Boost Collections and Recovery Results with Analytics.

related posts