One feature at a time

Pi-Squared had three game modes at launch. Two were technically fine and nobody played them. I cut them. Installs picked up almost immediately because the one mode left was actually good -- and the App Store listing finally said one clear thing instead of three blurry ones.

Reading about minimalism in product design did almost nothing for me. Cutting two features and seeing retention go up -- that built the dataset. Most product instincts are built that way: by deleting a thing you were proud of and being honest about the result a week later.

The original Stripe pitch was: drop seven lines of JavaScript into your site and you can accept payments. That was the whole demo. Patrick and John Collison stripped payments down to the one thing developers actually wanted to do -- and the discipline of that removal is still the brand.

I try to ask one question before adding anything new to a product, a page, or a calendar: what am I willing to take out to make room for it? If nothing, I usually do not add it. The constraint forces the trade-off out into the open.

Hit reply and tell me. I read every one.