Refund Program A/B Testing
research | UX | behavioral science
Federal Trade Commission
Role: UX designer and research catalyst
Context: The FTC provides money back to consumers after it sues companies for deceptive claims about products or services. When consumers can get a refund, the FTC usually mails people a check.
The Challenge: Identify design decisions to maximize response rates from the refund mailings.
I pitched the idea of a/b testing different variables to see what factors might influence check redemption. Having data to support design interventions would improve our decision-making. Even a null result would put speculation to rest and help us avoid costly interventions that didn’t actually drive higher response rates.
I worked with the Refund Administrator to get buy-in for the concept and validate feasibility. We designed a pilot study with 3 variables:
Envelope Color: White vs. Craft
Postage Type: Real Stamp vs. Bulk Mailing Indicia
Seal Color: Black vs. Color
Results: While I was at FTC, we completed 1 pilot run and the large-scale refund mailing was set up to continue the experiment. Partners were established within the FTC’s Redress Program, Bureau of Economics, and Claims Administrators.