On the last day of my 20s I wrote a line I keep coming back to: noise is extremely difficult to spot early on, optimize for prevention via structure, this compounds with time.
Four months into building solo I can name the noise concretely. A feature you shipped because somebody asked once and never used. A Slack channel you joined for a deal that didn't happen. A TODO comment from week two that you walk past every day. An unfinished branch on your laptop you keep meaning to delete. A contact you keep emailing because they once said "interesting." A side project you spend a Saturday on every month. A UI affordance for an edge case that hasn't happened in production yet.
Each one costs nothing on the day it's added. Each one costs a fixed small amount per week to maintain, ignore, or work around. Multiply by a year and you have a different company than the one you wanted. The killer property is that none of these are mistakes. Each was a defensible yes at the moment it was made. The damage is in the aggregate, not the instance.
You can't out-discipline noise. Willpower is expensive and renewable but limited. Structure is cheap and compounds.
A few things that worked for me. A deletion habit on Fridays for branches, files, tabs, todos, channels. Look for the second-cheapest thing to delete, delete it, the first-cheapest is too easy to count as work. One inbox rule per regret. Every time you regret reading something, write a rule that filters that category. Six months in the inbox is mostly signal. No half-shipped features in main, ship behind a flag or stash it, half-shipped lives in main and you forget which half. Kill the project that's almost ready. Almost is the strongest noise generator. A project that's been almost ready for two months is eating mental energy and producing zero.
The two-products mistake hit me hard. I spent the first three months running two products. They shared no code and no users. They didn't compete for the same hours but they competed for the same attention. The cost wasn't time. It was that every Monday morning I had to decide which one to work on.
I killed the second one in March. The remaining one moved twice as fast for the next six weeks. Not because I had more hours. The cleared mental space was the difference.
You almost never realize what the noise costs until you remove it. Then you do, and you remember why people who do one thing well so often beat the people doing three things adequately.
— Simon