tegtitle

I'd try rate limiting it before a full block. A Cloudflare rule that throttles that user agent keeps the site up, and you can still decide later whether you want to show up in their answers at all.

I'd only bother with the older posts that still get impressions in Search Console. Those get crawled often, so a link from them actually helps the new post get found, while adding links to posts nobody visits is mostly busywork.

Content first, but only stuff someone actually searches for. Ten articles nobody needs won't earn a single link either.

Search engines for me. The web was already there, but without something like PageRank most of it would have stayed unfindable, and pretty much everything else on the list depends on people actually finding stuff. Second place: version control.

Agree on the specific phrases part, for a small site that's pretty much the only way in. The big keywords are taken by sites that have been there for ten years.

Searching myself I get an AI answer for "what is X", but normal results for "X near me". Feels like it pushes small sites toward local and buying-intent pages and away from explainers.

For small niche sites in 2026 the biggest wins come from deep topical coverage topic clusters + real experience/insights that AI can’t just copy, clean technical stuff speed, mobile, indexing, Core Web Vitals, and a few relevant backlinks. Thin content + random links don’t cut it anymore.