Cyrus Shepard, founder of Zyppy SEO, went spelunking through 54 experiments, patents, and case studies and came back with a ranked list of 23 factors correlated with getting cited by AI search engines like ChatGPT, Gemini, and Perplexity.
The headline takeaway is unsexy on purpose: the top factors are URL accessibility, scoring 9.5, and plain old search rank, scoring 9.4. In other words, “AI visibility” does not seem to replace traditional SEO so much as check whether anyone did it properly.
Shepard’s scoring system weighs repeatability, strength of evidence — a 50-million-query study counts for more than a 10-query anecdote, mercifully — and “official support” from documentation, technical specifications, or patents. He also keeps the caveat the industry tends to misplace in a drawer: these are correlations, not proven causes.
Some of the more executional tells are where this gets interesting, and slightly annoying for anyone who loves a long intro. “Fan-out rank,” scoring 9.3, shows up near the top because AI answers are often built from multiple related searches, not just the original query. Ranking across a topic cluster, which scored 8.9, can therefore compound the odds of being pulled into an answer.
“Preview controls” scored 9.2, tied with “query-answer match.” That is a polite way of saying: if a publisher blocks snippets with directives like “nosnippet” or “data-nosnippet,” or uses AI-scraper defenses that limit visibility, it may also be blocking its own chances of being cited. A tidy little industry trap.
“Answer near the top,” at 8.8, is brutally practical. The analysis cites work by Dan Petrovic suggesting Google’s Gemini applies a retrieval cap per URL, which means content buried below the fold can miss the extraction window, regardless of how lovingly it was written.
The traffic numbers are messy in exactly the way this topic deserves. Shepard’s roundup cites Seer Interactive finding that appearing in Google’s AI Overviews correlates with 120% more organic clicks per impression and a 41% increase in paid clicks compared with when a brand is not cited. Another Seer analysis of 42 client organizations found 35% more organic clicks and 91% more paid clicks for cited brands, even as overall CTRs declined.
Meanwhile, Ahrefs research from February 2026 links AI Overviews with a 58% CTR reduction for top-ranking pages. SISTRIX puts Germany’s loss at 265 million organic clicks per month, with position-one CTR dropping from 27% to 11% when AI Overviews appear.
The contradiction is the point: AI features may siphon clicks broadly, while citations become a narrower lifeboat. One that apparently prefers accessible pages, extractable structure, and sentences that commit to verifiable specifics, which scored 8.3, rather than tasteful hedging. And for the record, Shepard found basically no credible evidence that LLMs.txt moves the needle, giving it a 2.0.
Full story at PPC Land.