Increasing Conversion for the DM Funnel (the $8 million test)

DM is the most successful channel for most credit card companies with the highest NPV, highest conversion, and also the highest CAC. Our goal was to take an experience with a lot of entry points, and simplify. Reduce emotional and cognitive friction for users and get them to their goal: application approval.

Increase direct mail application submissions.

In an effort to improve conversion for the direct mail funnel, we began ideating updates to the flow. The initial area of focus was the offer page, which was targeted after a deep dive into conversion metrics, Fullstory, heat maps, and scroll maps. This project proved to improve the business metrics by seeing an increase in conversion of 8%, an improvement which correlates to an additional million dollars per year, per percent improved (again, for everyone in the back, that’s an $8 million a year improvement).

The fun won’t stop here! This test was the just the MVP and I am currently working through next steps to bring ideal state to test, then hopefully to market.

During discovery, we noted that 92% of users that landed on the offer page (left) immediately clicked through to apply now. This demonstrated a high intent audience which begged the question: why were we making applying harder?

While there were internal thoughts on what content users should see, the reality was the large majority were choosing not to see it and moving forward to apply.

For this update, I needed to ensure that users were still able to see the terms of their offer and any other requirements for disclosures.

Here the form is visible in the first load before scrolling so that users looking to get started right away are able to. I also changed the rates and terms to a pop up instead of anchoring to the bottom of the page. This would mean that users could either click away from the pop up or click to close and easily continue their application.

Overall, we were able to get this test out the door quickly, making several choices that wouldn’t harm the test, but would reduce time to build. This allowed in market validation of a core suspicion that we have several places in the funnel with harmful friction.