Research library Customer value, acquisition and retention
Don't Chase the Customers Most Likely to Leave.
Eva Ascarza ran retention offers in two field experiments. Aiming them at the customers the offer would actually change beat aiming them at the customers most at risk.
Eva Ascarza, 2018
Extra drop in churn from targeting by response instead of by risk
What she was looking at
Most retention programs run the same way. A model scores every customer’s risk of leaving, and the save offer goes to the highest scores.
Eva Ascarza, a marketing professor, tested that rule in two field experiments:
- A prepaid wireless provider in the Middle East. 12,137 eligible customers were randomly split, and the treated group got a text offering free credit when they recharged.
- A North American membership organization. Some renewal letters included a thank-you gift.
She then used machine learning to estimate how much each customer responded to the offer, and compared two ways of choosing whom to target: by risk of leaving, or by predicted response to the offer.
What she found
“customers identified as having the highest risk of churning are not necessarily the best targets for proactive churn programs.” Ascarza (2018), abstract, p. 80
Targeting by response cut churn by an extra 4.1 percentage points in the first study and 8.7 in the second, with the same campaign. The customers most likely to leave and the customers the offer would actually change overlapped by only about half.
“retention programs are sometimes futile not because firms offer the wrong incentives but because they do not apply the right targeting rules.” Same
The paper won the American Marketing Association’s Paul E. Green Award for 2018.
What it means
Risk and response are different questions. Some high-risk customers are leaving whatever you offer. Some low-risk customers would have stayed anyway. Retention money only earns a return where it changes the outcome.
Related work makes it sharper. In one experiment, proactive retention outreach raised churn: 10% of the contacted customers left within three months, against 6% of those left alone.
Where it shows up in a revenue team
- Save campaigns aimed at the riskiest accounts spend heavily on lost causes.
- Weight by value, too. The right target is the customer the effort changes, weighted by what that customer is worth.
- Keep a holdout. Leave a random slice untreated, so you can see whom the offer actually moved, not just who stayed.