We’ve been working on a visual refresh of Buffer’s web app for a while now: new navigation, updated design language, the works 💅

I think a lot of my teammates (including myself, at various points) assumed this was mostly cosmetic, without much in terms of concrete impact.

So we A/B tested it with new users, and virtually every key metric hit statistical significance in 4 days 🤩

The variant with the new design showed a meaningful lift in activation rate itself and each one of its leading indicators: new users connecting channels, creating their first post, replying to comments, actually getting started. Even short-term conversion to paid plans improved!

I keep coming back to this because it challenges a bias I’ve held at different points in my career: that design refreshes are really hard to justify with data because the impact is diffuse and slow.

Sometimes that’s true. But sometimes a clearer, more modern interface just removes friction to such a degree that it shows up in key growth metrics. People see a product that looks current, calmer, a bit easier to navigate, and they trust it a little more. Folks keep going 🚀

We’re rolling it out broadly this week. I’m curious how the numbers hold as the sample grows and the novelty fades! But even as a directional signal, it’s a good reminder that design is product work, not decoration.