Why I started

I saw a reel claiming a timer site makes $650K a year. People find it through search, ads pay, the owner barely touches it. It looked like a perfect fit for a $100-a-day experiment.

I had the AI find 12 candidates using only free methods: Google autocomplete to see if people actually search for it, the first results page to see who is already there, and Trends to compare volume. Workout timers had the best balance of demand and competition. The first page for "boxing timer online" was all small sites.

What one day produced

All on day one: the timer, pages per mode, the domain (workoutbell.com, $10.44 a year), deployment, Search Console, the AdSense application, a gym-wall-clock design, Portuguese, Korean and Italian pages, and GA4. 24 pages in total.

Three days later I fixed speed. Mobile score was 59 and the layout jumped a lot. After the fix it scored 100.

Where it got stuck

AdSense said no: "low value content". There were many pages, but each had 90–350 words, and the preset and translated pages shared almost the same template. Making a lot of pages fast became the weakness.

I wrote 14 new workout guides, fact-checked them and requested indexing. On October 18 I hit the re-review button.

What a human had to do

Pay for the domain, sign up for AdSense, click the email verification, press the re-review button. Anywhere money moved or my name was on the line, a person had to do it.

What I learned

Building took 3.6 hours. The bottleneck wasn't building, it was waiting. Search traffic takes months and ad review is a human decision. The AI can be fast, Google's clock doesn't speed up.