Challenges in redesigning a site with 1.5 million users

Take an outdated site with boat loads of users and redesign it. Sounds great! But it's not as simple as just re-skinning old with the new. There are thousands of daily active users who would be confused at an immediate rollout of a new design. Not to mention you could cost the company 💵 from a drop in sales!

How then?

Bit by bit, is how. Slowly rolling out test after test and improvement after improvement. A/B testing, failing, trying again, testing, repeat. I actually found it incredibly hard to maintain my focus on the big picture while redesigning a small portion of the checkout system, for instance. It's a slow process and requires loads of patience, or it can get frustrating. You test what works, what improvements are worth doing, and what aren't.

Sometimes it's best to design improvements using the same style currently in place, in preparation for switching it out with new UI later. It's a way of working that I hadn't really experienced before, but I loved it. Conducting user tests and interviews, prototyping — it was UX at its core. I learned a whole new way of thinking about design. Practicality first, solving the problems that presenting before even thinking about how it could look.


With the slow pace of things I was able to properly document new design patterns for future designers. This would leave no ambiguity in what components were meant for what. Example:

In the background I was creating a new design system based on improvements we were rolling out in the current style. Untill eventually it was time to start mixing in the new designs. It's messy, and for a while, you are left with a mix of old and new components. Patience again, we only rolled out what made sense first.

Old and New