AI search & GEO

Getting cited by ChatGPT, Gemini, Perplexity and AI Overviews — and how that differs from ranking.

14 articles

SEO vs GEO: is generative engine optimisation actually a different job?
A sceptical, practical comparison of SEO and generative engine optimisation — what genuinely differs when the destination is an AI answer rather than a ranked list, and what is rebranding.
5 min read
SEO vs AEO: answer engine optimisation, minus the hype
What answer engine optimisation means in practice, how it differs from ordinary SEO, and which parts of it are genuinely worth changing your process for.
4 min read
GEO vs AEO vs SEO: three acronyms, one job, honestly separated
What generative engine optimisation, answer engine optimisation and search engine optimisation actually mean, where they genuinely differ, and how much of the distinction is marketing.
4 min read
Traditional SEO vs AI-era SEO: what actually changed
An honest audit of which SEO practices still hold in the age of AI answers, which have quietly stopped working, and which never worked in the first place.
4 min read
Ranking vs being cited: two different outcomes, two different playbooks
Ranking in results and being cited in an AI answer are different outcomes with different mechanics and different business value. How to tell which one your business actually needs.
5 min read
AI Overviews vs organic results: what changes below the answer
How AI-generated summaries at the top of search results change the value of ranking underneath them, which query types are most affected, and what to do about it.
4 min read
ChatGPT vs Google for discovery: where people actually find things now
How assistant-led discovery differs from search-led discovery, what it means for your funnel and attribution, and which one you should be optimising for.
4 min read
How AI assistants pick sources — what is known, and what is guesswork
What can actually be established about how AI assistants choose which sites to cite, what is reasonable inference, and what is confident speculation sold as fact.
4 min read
llms.txt vs robots.txt: one is a standard, the other is a proposal
What robots.txt does, what llms.txt proposes to do, which one search engines and AI crawlers actually honour, and how to control AI crawler access properly.
3 min read
Schema markup for AI vs Google: same markup, different payoff
Whether structured data helps AI assistants the way it helps Google, which schema types are worth implementing, and how to avoid the markup that does nothing.
4 min read
Brand mentions vs backlinks in AI search: why the unlinked mention matters more now
How unlinked brand mentions and traditional backlinks differ in an era of AI answers, and why PR without links is no longer the consolation prize it used to be.
3 min read
AI visibility tools compared: what they measure and what they cannot
Peec AI, Profound, Semrush and Surfer all track AI-assistant visibility. What each actually measures, how reliable the methodology is, and whether you need one yet.
4 min read
Content for humans vs content for LLMs: is there actually a difference?
Whether writing for AI systems requires anything different from writing well for people, what the genuine differences are, and the advice in this space worth ignoring.
4 min read
Perplexity vs Google traffic: what the referral numbers actually mean
How traffic from answer engines like Perplexity compares to Google organic in volume and quality, and how to measure it without fooling yourself.
3 min read

RankBrain AI

Is RankBrain AI a good fit for your workflow?

If you have findings to implement and access to your site’s code, an editor-based workflow may help you turn those findings into changes you can review. If your main need is backlink research, historical rank tracking or client reporting, a dedicated data tool may be a better fit.

RankBrain AI connects SEO guidance to Claude Code, Cursor or Codex in your editor. With the relevant file access and tools configured, your assistant can inspect templates and propose code changes. You review the diff, test it and decide what to ship. This is implementation support, not a guarantee of rankings or a replacement for specialist datasets.

A good fit: you use an AI coding assistant and want SEO guidance alongside the files you maintain. Not a good fit: you need a managed service or a standalone backlink and historical ranking database.

Do it yourself — free

You own the site and the repo, so you already have what you need. Point your AI coding tool at the codebase, work through the fixes in this article, and pay nobody. Most of what is written here is method, and the method is free.

Consider RankBrain AI — $5/month

The hosted corpus gives your assistant a shared reference for SEO implementation. Check its advice against current documentation and your own tests. Flat $5/month, no seats, 3-day trial. Use the trial to see whether it helps with a real task in your workflow.

Check my site free See how it works

We make this, so weigh the comparisons accordingly — and check them against the capability tables, where we sit on the same rubric as everyone else. We have no backlink index and no historical rank data, and we say so there.