In February, I decided to build five content sites from scratch and get them to 10,000 words of published content each. Deadline: end of February.

I did it. Just barely. Here's what I learned.

The setup

Each site targeted a narrow niche: 1. RV Composting Toilets — Educational content about composting toilet options 2. Portable Solar Power — Buyers guides for portable solar panels 3. Small Space Living — Design and productivity tips for small apartments 4. Niche Hobby #1 — Very specific hobby community 5. Niche Hobby #2 — Another specific community

Zero budget beyond the AI tools I already pay for. Goals:

  • 10 posts per site (10,000 words total)
  • Domain + hosting
  • SEO basic optimization
  • No promotion
  • Timeline: Feb 1 to Feb 28. One month.

    What worked

    The content pipeline I built actually worked at scale.

    Remember the content engine from my earlier post? That system generated 50 drafts in the first two weeks. I tweaked 40 of them and published 35.

    That system was designed for this. Generate fast. Edit for specificity. Publish.

    Niche selection was crucial.

    RV Composting Toilets is narrow. Very narrow. But that narrowness meant:

  • Easy to find competitors (only 3)
  • Easy to rank (no big sites competing)
  • Easy to find your angle (theirs were all terrible)
  • Broader niches didn't work. Small Space Living got 15,000 words written but only published 8 posts because I couldn't find a unique angle. Too much competition.

    AI content + manual editing works at scale.

    I generated drafts in Gemini (cheap), edited them in Claude (better), published in 3 hours for 10 posts.

    The content isn't perfect. But it's good enough to rank. And I've tested it — 30 days in, the RV Composting site already has three posts ranking in top 10.

    Automation for hosting and publishing saved time.