Week 8: The Next Build Is a Directory. Which One? Help Me Pick.

It is Monday of week 8, and for the first time in this project, the build is starting before the idea is locked. This week's build is a database or directory play. Which niche? Genuinely undecided. Two candidates are on the whiteboard, one wildcard came in through the idea box, and the pick is happening in public this week. But first, the week that just ended deserves its paragraph, because it was the biggest one yet.

Week 7: the week everything landed

Three things happened last week that this project had never seen. The first money arrived: $33.81 in affiliate commissions from AI Gift Pro, seven weeks of zeros finally broken. The week 7 build shipped on time, Friday July 17 as promised: it was announced as an ADD toolkit and collapsed, deliberately, into its sharpest tool, RSD Decoder, and after the AI Gift Pro slip that on-time Friday mattered more than anything. And the portfolio grew sideways: two side builds went live the same weekend, PitchLite and Medicare Decoder. Eleven builds live. Revenue: $33.81. Momentum: real.

Why a directory, and why now

Because the quiet pattern in my own numbers points there. ThrowSpot mapped 7,008 disc golf courses and keeps ticking with zero promotion. BJJMath, structured calculators built from real rulebooks, pulled the portfolio's best early organic traffic. Meanwhile the generic-article internet is being eaten alive by AI answer boxes, and the assets that survive are exactly what directories are: comprehensive, structured, interactive data that a summary cannot replace. The 2026 playbook I keep writing about in the journal says tools and data beat articles. Time to eat my own cooking, again, at a bigger scale.

The criteria for the pick

Four filters, and the niche has to pass all of them. A passionate audience that actually searches: hobbyists, professionals, or a life moment with urgency. A public data source that can be structured: government records, published rulebooks, open datasets, something real to build on rather than content to invent. A weak incumbent: the niche either has no good directory or has one that looks like 2011. And a believable money path: featured listings, affiliate handoffs, or lead generation that can be disclosed honestly, the Medicare Decoder way. ThrowSpot passed all four by accident. This one passes them on purpose.

The open question

I am not naming the two whiteboard candidates yet, partly because the wildcard from the idea box is genuinely competitive and I want to sit with it until Wednesday. What I will say: one candidate is a hobby niche, one is a professional niche, and the wildcard is a life-moment niche. The pick gets locked midweek, the build ships Friday, and the full reasoning, including the two ideas that lost, goes in next Monday's log. If you have a niche that deserves the ThrowSpot treatment, the idea box is open through Wednesday and I read everything.

The other half of the week

Per the distribution plan, Thursday and Friday are now promotion days, and this week they have a clear job: get the three side builds in front of their communities, start the seasonal SEO groundwork for AI Gift Pro before gifting season, and keep showing up where the twelve builds' people actually gather. The scoreboard is the public revenue page, which finally has a number on it worth beating. Numbers as of this Monday: $33.81 earned, eleven builds live, 12 followers, 0 subscribers, and one niche to pick. See you Friday.

Update, since this post went up: TextScrubr sold the first copy of its Claude skill, $29, the project's first product revenue, and the Amazon affiliate total climbed to $50. Running total: $79, all on the revenue page.