Want to get more out of your recovery scoring strategy? Create a control group.
You did it. You got your company on board with using a data-driven recovery strategy and you’re about to implement it. You’re also going to be held accountable for results. Wouldn’t it be helpful to:
If you answered yes to any of these questions, consider creating a control group.
A control group allows you to know whether or not your scoring strategy is working, and can provide insights into where to make necessary improvements. It’s imperative to truly understand the value recovery scores bring to your business and to prove a return on investment.
A scientific approach to creating a control group is to gather a random sample of accounts, equal across each alpha group, totaling 2-5% of your overall inventory. Then, treat all those accounts with the maximum treatment strategy you’ve assigned to your A group.
Even with the maximum treatment, within a control group, A accounts should still perform like A accounts, B accounts should perform like B accounts, and so on. In essence, you’re able to verify that even if you treat your D accounts with the maximum A group strategy, they still behave like D accounts.
In my last post on developing collections strategy constants, I proposed an optimal structure of strategy constants in which the D group accounts receive the minimum standard of work.
The idea is that D accounts do not perform as well, and as such it’s not a wise investment to over allocate data and resources towards this underperforming group. Your investment is much better placed in your higher performing alpha groups. In fact, we’ve seen cases in which 47% of money recovered came from the top 10% of inventory—alpha group A. (Learn more about a typical sloped data acquisition plan in step 6 of our collections strategy insight guide.)
A typical sloped data strategy
|Priority level||% of $ recovered*||Typical data strategy|
Top 10% of scored accounts
|47%||Address, phone, triggers, POE, VPOE, manual skip tools, property, assets, relatives|
||36%||Address, phone, triggers, POE, VPOE, manual skip tools|
|13%||Address, phone, triggers, POE|
|5%||Address, phone, triggers|
*Based on CreditVision Recovery Scores
If you’re following a sloped plan and putting the least amount of collections effort and expense toward recovering D accounts, it’s reasonable to get a minimal return. However, when you use a sophisticated recovery score like TransUnion’s CreditVision recovery scores, the predictability of the data—validated by your control group—shows that you can put any amount of effort and expense toward D group accounts yet you’ll still only recover a minimal amount.
If you don’t perform a control, you won’t know if the score is correctly predicting that D accounts will underperform, or if it’s the treatment that’s causing the D accounts to underperform. In other words, your D group strategy becomes a self-fulfilling prophecy that makes you question the value you’re getting out of scores.
TransUnion "If you don’t have a control group, you can't measure reliably. You don't know if it's the score of the treatment that's causing a particular behavior and your D group strategy becomes a self-fulfilling prophecy that makes you question the value you're getting out of scores."
Don’t skip this integral step in your debt recovery strategy. You want to show you made the smart move for your business by taking a data-driven approach, so let the data work for you. Create a control group to prove your scores and strategy are working as they should.
If something doesn’t look right, you’ll have the insights to change things before too much money and resources are allocated to the wrong accounts.
Want to learn more about building an effective, data-driven recovery approach to optimize your collections process? Read our insight guide for 10 Steps to Drive More Profitable Collections Operations.
Strike the Right Balance Between Fighting Fraud and CX: Four Findings
TransUnion delivers top-notch service to any sized business
Balancing Fraud and the Customer Experience as Fraud Evolves
2019 Predictions: Consumer Credit, Balance and Delinquency Rates