Premium improvement

Improving Premium conversions

Locked content, before and after.

Locked content, before and after.


My Contributions

  • As the team’s product manager, I identified and led a project that increased payment page click-throughs by 55%

  • Collaborated with the data team to validate assumptions about web conversion problem areas

  • Worked closely with design and engineering to create an lean, iterative solution

Problem

How might we increase conversions for qualified users on web?

Geocaching relies on a Premium subscription model that offers additional content and features. The Premium features are detailed on an upsell page at the beginning of the payment flow. Data showed that click-through rates to the upsell page were extremely low–especially when compared to a similar page on the native mobile apps.


Process

The team’s designer ran a desirability test on the existing screen. User described the page as "untrustworthy", "ugly", "confusing", "combative", and "restrictive". One memorable quote became a rallying point for the team to fix this page: "You have to click to get any sort of information. No other word on this list was negative enough to express my disappointment." This qualitative information signaled that we were on the right track and that the poor conversation rate was a result of an unclear and off-brand message.

The designer tested a few different layouts and messages. We considered a multi-variant test with a matrix of swappable layouts, CTAs, and messages. I made a call to keep it simple and learn as much as possible while keeping the scope small. We moved forward with a test of two variants and the existing design as a control.

 
body_paywall04.png
 

Response and future considerations

After three weeks of A/B/C testing, Google Optimize showed that the "padlock" variant won with statistical significance. During this testing time period, the padlock variant increased clickthroughs to the payments page by 55% and is predicted to convert 1,000 more premium members annually.

Now that a scaffolding exists for testing different layouts and copy on this page, the team can continue to iterative to look for additional clickthrough improvements.

 
body_paywall05.png