SEO Recap covering August 7, 2026: cats.txt satire and the case against llms.txt

Built by Stephanie Chung·4 min read·5 stories

Daily summary of what matters in SEO, GEO, AEO, and AI search generated with Claude Code (beware of hallucinations)

Generative Engine Optimization (GEO) & Answer Engine Optimization (AEO)

Winning in AI search relies on the same fundamentals Google has pushed for a decade: clarity, a healthy site and off-site presence, and letting the business drive SEO rather than chasing tactics.
Jono Alderson, Mordy Oberstein, and Ross Hudgens each argued within days that the SEO vs GEO naming fight is a distraction from the real work.
The genuinely new emphasis is off-site entity optimization and consistency across every place a brand appears, because a model assembles its picture from everywhere it can read, not just the homepage.
Instead of testing category prompts like "best CRM for small teams," prompt LLMs and AI Overviews about your brand to see what they believe and which sources they pull from, then fix those exact pages and profiles.

YouGov BrandIndex found Claude's Consideration among US Gen Z adults nearly doubled in Q2 2026, from 14.2% to 28.1%, the highest of any brand in the ranking. OpenAI rose from 10.1% to 22.1% and ChatGPT grew more than 50%, from 16.3% to 25.4%.
A separate 19-market YouGov survey put US trust in AI assistants at just 28%, 42 points behind trust in traditional search engines, showing consideration and trust are different currencies.
Consideration this quarter moved on paid media and seasonal promotions (Johnnie Walker, REI) rather than GEO tactics, a reminder that broad-channel brand building still drives consumer consideration.
Track AI-platform awareness and consideration separately from whether your content gets cited, and build PR around named sources, verifiable data, and transparent methodology to close the trust gap.

To test the standard of proof behind llms.txt, the author invented cats.txt, a file declaring office cats with names, jobs, breeds, and a PurrLevel score, and it passed all four "proofs" cited for llms.txt.
The four flawed proofs are debunked one by one:
- AI bots crawling the file just means crawlers fetch everything, not that contents are read or trusted
- Google indexing it only confirms a URL exists with words on it
- An LLM repeating a file-only fact is ordinary retrieval-augmented generation treating it as any web page
- ChatGPT endorsing it reflects the average of online discourse, not an assessment
John Mueller said no AI system currently uses llms.txt and that consumer LLMs fetch pages for training and grounding but not the llms.txt file.
Ahrefs analyzed 100,000 domains and found the file largely ignored by crawlers, with other studies showing no measurable citation advantage for sites that add one.
The "convergence problem" is that a model asked whether llms.txt works returns the most common thing written about it, not reasoning. Once the community admitted cats.txt was a joke, the model shifted to calling it satire.

AI crawlers now hit sites thousands of times while sending little or no traffic, and Conductor's Search Console data shows impressions rising while clicks fall, so reporting should shift away from organic sessions.
The average AI prompt runs 23 words versus three to four for a Google keyword, so Conductor builds a scored custom prompt index rather than a keyword list.
YouTube is the number one cited site across all major LLMs, and Reddit citations signal the model found no better source, marking a content gap a brand can fill.
Pat Reinhart said backlinks retain traditional-search value but carry little weight for AI visibility, and advised against changing publish dates just to look fresh.

Gary Illyes said Google's crawlers send HTTP methods beyond GET and POST, including HEAD, OPTIONS, PUT, PATCH, and DELETE.
Those additional methods make up less than 1.5% of total crawler requests and are triggered by JavaScript during rendering.

Published on August 8, 2026