The lab — Day 2 of 365 —
We threw away half the keywords
The internet is full of people who will sell you a cadence. Two posts a week. Sixteen a month. Publish daily or Google forgets you exist. We almost wrote that into the plan as if it were physics.
It isn’t. Google ranks URLs, one at a time, and ignores thin copies of the same question. There is no counter somewhere awarding points for a calendar. The fake law was covering for something true and much smaller: don’t ship seven pages that are the same page in a different hat. The real ceiling on how much you can publish is how many genuinely different things people are searching for. For us that number is small, and knowing it is small is more useful than a posting schedule.
3,185 keywords in, 1,761 out
We pulled three Keyword Planner exports: one seeded from Daylio’s site, one from Finch’s, and one giant “overall” plan that came back containing baseball teams and medical journals. Union of all three: 3,185 rows. We kept 1,761 and deleted the rest.
Filtering is where the judgement lives, so it is worth being specific about what got cut. Anything that was a different product wearing our words. Anything clinical — we are not going to write a bipolar mood tracker page to catch a low-competition number, because we do not make a medical device and the person typing that deserves better than our SEO. Anything about a year we are not in. Junk journal, which is a beautiful craft and not this app.
What survived sorted into clusters, and the sizes are instructive:
| Cluster | Keywords | What we did |
|---|---|---|
| Paper bujo | 918 | Nothing. Biggest pile, least ours. |
| Journal / diary | 415 | Nothing. Head terms owned by apps with a decade on us. |
| Mood tracker | 199 | One page. Not a war with Daylio. |
| Digital bujo | 116 | Started here. |
| Year in Pixels | 41 | Waiting. It’s ours, and it can wait. |
The largest cluster got zero pages. That is the part that feels wrong and isn’t. 918 keywords about paper spreads is 918 keywords belonging to people who want a notebook, and we do not sell one.
Why we started where the store already wins
The Play Store listing is already doing well on digital bullet journal. So the site started there too — same phrase, two front doors — rather than opening with a fantasy about outranking Daylio for mood tracker on a domain that barely exists yet.
The first essay went live: the notebook wasn’t in the room. One idea, one URL, no spec box on the page explaining to the reader that they are inside a content strategy.
We stopped advertising a door that isn’t open
Every TestFlight mention came off the live site. There was no iOS build to let anyone into. A beta badge for a beta that does not exist is a small lie that costs you the one thing an indie app has, which is that people believe you.
The rest of the day was unglamorous: page-speed failures from two audit tools, CSS that blocked first paint, media without reserved slots so the layout jumped, and meta descriptions on the legal pages that were too short to be a sentence. Also an email that actually sends people the Play Store link when they ask for early access, instead of collecting an address and doing nothing with it.