I tracked every hour I spent on my business goals for almost 3 years. 56% of it was procrastination that looked like work.

I log my time and every task I work on gets a timer. I started in 2021 so there is a lot of data by now.

I just pulled everything I logged against an actual goal between January 2024 and this month. It’s 4,393 hours. Then I went through all 740 tasks one by one and asked the same question about each one. Did this actually move the goal or did it just feel like it did?

56% did not, which is 2,469 hours. About 14 months of full time work.

Per year 2024 was 73% wasted, 2025 was 60% and 2026 so far is 40%. It is getting better but 2024 is brutal to look at.

Almost none of it was obvious laziness. I was working hard on things that were close to the goal but did not move it forward.

The two worst goals were both business goals. One was making $5,000 a month from a product I was selling, 1,192 hours and 66% of that was fake, and I eventually abandoned it. The other was building an agency and getting my first clients, 298 hours, 88% fake.

So what it looked like is I spent 42 hours writing a procedure for preparing client proposals, and I had no clients. Then 40 hours on a procedure for responding to leads in my pipeline, 26 hours on a lead outreach procedure instead of doing outreach, and 37 hours building a process for tracking metrics I did not have yet.

There were also 109 hours on a content production process for YouTube, so I was building the process for recording videos instead of recording them. And 51 hours on an offer strategy.

9 of my 15 biggest wasted tasks start with the words “I created a process” or “I created a procedure” or “I created a system.”

The worst version was in how I wrote the goals themselves. One goal was to get 5 beta testers for something I was building, and the definition of done I wrote for it had five separate systems “fully finished” plus a voice agent. Before a single person had tested anything. So every hour I spent building looked like progress, because I had written the building into the goal.

Goals with a physical result were clean. Organizing my move to Spain was 98% real work, because there is nothing to systematize when the goal is “live in Spain.” The silliest thing I found was 19.6 hours sorting my Spotify library by genre, logged under a goal to play a 30 minute DJ set for at least 10 people.

Goals that needed other people to say yes were the dirty ones. Writing a procedure is fully under my control but a stranger deciding to pay me is not, so I kept picking the part I controlled and called it work.

So what I do now is before a task goes under a goal I ask whether anyone outside my computer could tell the difference if I skipped it. If I skip “write the outreach procedure” nothing changes for anyone, and if I skip “send the messages” something does. The first one goes on a someday list and the second one gets a timer. I also stopped writing building into my definitions of done, so now it is something that happens outside my own screen, like someone paid or someone used it.

I don’t trust myself to catch this while I am doing it because the whole point is that it feels like work at the time. So I set up my AI system to challenge me when I add a task like that and ask which result it moves. That system is Execution Compass, the planning app I’m building, and all the numbers above come from my own data in it.

I am at 40% fake this year instead of 73%. Still not great.