Point.me had built something really impressive under the hood: a search engine that crawled hundreds of sites and flight APIs to calculate points value, transfers, and pricing. But the product itself wasn't converting people, and it wasn't keeping them around either.
I led a redesign that touched three parts of the experience: the information architecture, the free tier, and the booking flow, all aimed at making the product actually work for people who weren't already points experts.
The technology behind Point.me was solid. The experience wrapped around it wasn't.
The current Information Architecture (IA) was quietly killing signups before people ever got a real look at the product. You'd land on the homepage, hit a wall almost immediately, and end up on a pricing page that didn't tell you much of anything. There was also nothing in the free tier worth sticking around for. If someone wasn't ready to book right that second, there was no reason to come back tomorrow and poke around, so we were losing the exact users who needed time to warm up before they'd ever pay for anything. And the booking flow itself was built like it assumed you already knew what you were doing. Dense screens full of information made total sense if you were a seasoned points traveler, and completely lost anyone who wasn't.
None of these problems existed on their own. A confusing IA meant fewer people got far enough in to see what the product could do. A thin free experience meant the ones who did get in had no reason to stay. And a booking flow built for experts meant even the people who made it that far could still fall off right before converting.
shifting focus from expert travelers (a shrinking, contested segment) to novices (an underserved, larger one). Triaging users from their first session, instead of forcing every skill level through one generic flow
Explore gave people real reasons to stay engaged, which made the eventual upgrade path feel earned rather than forced
— every change was measured, which is what actually de-risked leadership's biggest fear
— the concierge escape hatch meant users never felt stuck, even in a genuinely confusing domain