Stephen Sumner
The 120-Point AI Readiness Framework: How I Actually Audit Sites for AI Search
The full rubric I run on every client site: what each category measures, why it is weighted the way it is, and what I deliberately refuse to score.
4 June 2026· 18 min
Long-form writing on AI search, local SEO and the legal sector: the frameworks I build, the things I test, and the notes I write to think problems through.
The llms.txt proposal was revised on 10 August 2026, and some of what you were told about it is now out of date. What changed in v2, the honest read on whether it earns its keep, and a checklist for getting yours right.
The full rubric I run on every client site: what each category measures, why it is weighted the way it is, and what I deliberately refuse to score.
The handful of changes that reliably shift a page from indexed-but-ignored to quoted: top-loading, schema honesty, and the snippet directives nobody remembers setting.
Why NAP consistency still decides local rankings, and a boring, repeatable process for fixing years of drift across directories.
Practice-area pages, review strategy and the trust signals that matter when the searcher has a problem and very little patience.