Selection Rate Optimisation: What Actually Wins the Retrieval Gate in AI Search
What wins before the AI writes its answer? 385 passages, 30 queries, 120 judge calls on Gemini and ChatGPT. Entity naming lifts 2.7–4.2x. Data and playbook.
Read post01 / Blog
The Blog
Thoughts on Search, AI adaptation, systems thinking, general marketing and building search visibility that works beyond just Google.
What AI Actually Cites From Long-Form Pages
I ran 480 API calls across ChatGPT, Gemini, Claude and Perplexity to find where AI really cites from long pages. Original data plus how to structure content for it.
Read postWhy Claude Penalises the Exact Content ChatGPT Rewards
Claude can suppress the same brand page ChatGPT rewards. I ran 540 tests to prove it. Free Injection Risk Scorer scores your copy on both axes.
Read postGrounding vs. In-Model: The Query Classifier Every SEO Needs
Before fan-out or SRO work, ask whether your query is even retrieval-sensitive. Learn the grounded vs in-model split and classify queries before you optimise.
Read postSelection Rate Optimisation: The AI SEO Metric Replacing CTR
Learn what Selection Rate Optimisation is, why CTR is fading in AI search, and how to improve your chances of being cited in AI Overviews.
Read postFan-Out Query Reverse Engineering
AI search doesn't use one query. See how ChatGPT and Gemini generate fan-out queries, why keyword tools miss them, and how to adapt your content.
Read postScreaming Frog v24 MCP: The Complete Agency Guide. What It Does, What It Doesn't, and the Prompts That Actually Work
A practical agency guide to Screaming Frog's official MCP server, where it helps, where it falls short, and the prompt chains that make it useful.
Read post