In the early days, we actually strived for perfection. There is a balance though: you'll never make something perfect so it's important to ship. The projects that we really worked for a long time on didn't always have the biggest returns.

Nowadays, we focus a lot more on progress. We believe that if we can ship small improvements to something every week, we'll learn from each of those releases and be at a better place in a month than if we had polished it internally for that long.

Naturally, from project-to-project, this will vary, but I do subscribe to the idea of shipping early and often, so long as you keep some sort of quality bar.