When the best work is subtraction

I spent most of yesterday taking things out.

Not fixing broken things. Not adding new capabilities. Just removing. A module nobody had asked for since it was built. A permissions matrix that turned a simple action into a small bureaucratic exercise. A cluster of folders that had migrated over from an older system and were sitting there — furniture nobody agreed to buy, but nobody wanted to be the one to move.

By the end of the day the product had fewer options, fewer tabs, fewer settings. It did less on paper. It felt more useful in practice.

This is hard to internalize when you're building something alone. There's a deep pull toward addition. Adding a feature is legible work — you can point to it, demo it, describe it in a line. Removing a feature is also work, but it feels like confession. Like admitting the thing was never quite right, and now you have to own that in front of yourself.

Entrepreneurs are especially susceptible to this. The loop we live in already makes us feel behind — building the product before the customers exist, finding the customers before the product is ready. In that headspace, removal feels like going backwards. But it isn't. It's often the only honest way forward.

The product starts to carry its own history

At some point, everything you've shipped accumulates like ballast. Every feature built for a customer who was imagined rather than real. Every configuration option added because you weren't confident enough to just decide. Every good idea from six months ago that turned out to be a medium idea when someone actually tried to use it.

You can keep sailing with all of it on board. Most people do. But the boat gets slower, and the drag is invisible until you strip something out and suddenly it's not.

Patton's line — a good plan violently executed now beats a perfect plan next week — is true for new territory, when paralysis is the real enemy. But for something that's been running awhile, the violent execution sometimes needs to go the other direction. Not forward into more. Backward into less.

Deletion is a decision. It's actually a sharper decision than addition, because addition says maybe this. Deletion says not this — and means it.

The product that gets lighter is usually the product that gets used.