You rank on Google.AI has never heard of you.
AI is the new search. Your buyers now ask ChatGPT, Gemini and Claude for a recommendation and take the answer they get. I make sure your brand is in that answer — without giving up the Google rankings you already paid for.
Ranking and being recommended are different jobs
Traditional SEO still matters — AI engines lean heavily on what already ranks. But a good SEO programme was built to win a results page, and a growing share of buyers never see one. Answer engine optimization (AEO) and generative engine optimization (GEO) are the work of getting chosen when the machine writes the answer.
Built for brands that already did SEO right
If you have no SEO foundation, start there. This service is for companies with traffic, content and rankings who have noticed the leads are asking about a competitor they heard about from an AI.
Page one on Google, absent from the answer
AI describes you wrong
The crawlers cannot see your site
Three layers an AI checks before it names you
A model recommends a brand when it can read the site, extract a clear answer from it, and find the rest of the web agreeing. Miss any one layer and the other two stop counting. I am an engineer first, so the technical layer is never the part that gets skipped.
Can it read you?
Technical SEO, extended to the crawlers behind AI answers.
- →Server-rendered pages so crawlers read content without running JavaScript
- →robots.txt and firewall rules that let the AI crawlers you want through
- →Schema.org markup for your organization, products, people, FAQs and reviews
- →An llms.txt and clean, crawlable sitemaps pointing at what matters
- →Core Web Vitals, canonicals and internal linking kept healthy for Google
- →Facts in HTML text — not trapped in images, PDFs or tabs that never load
Can it quote you?
Content shaped so a model can lift a passage and trust it.
- →Pages that answer the exact questions buyers ask assistants, up front
- →Passages that stand on their own when lifted out of the page
- →Comparison and “best for” pages that make your positioning quotable
- →Original data, pricing and specifics a model can cite instead of guess
- →Named, credentialed authors and dates that signal the page is current
- →Your existing top-ranking content restructured rather than rewritten
Does the web agree?
The off-site signals that decide whether a model believes you.
- →One consistent description of what you do, everywhere it appears
- →Knowledge panels, Wikidata and business profiles corrected and filled in
- →Presence on the review sites, directories and lists that models cite
- →Earned mentions in the publications and communities your market reads
- →Reddit, YouTube and forum answers where the real conversations happen
- →Outdated facts about you tracked down and fixed at the source
A baseline first, so every result is provable
Nobody can promise a model will recommend you. What I can do is measure where you stand, fix what is in the way, and show you the change on the same questions every month.
- Weeks 1–2
Measure where you stand
I build a set of the prompts your buyers actually use — category questions, comparisons, “best tool for” — and run them across every major AI platform. You get your share of answers, who is named instead of you, and which sources the models cite.
- Weeks 2–4
Fix what blocks the crawlers
A technical audit of your site as an AI crawler sees it: rendering, bot access, structured data, sitemaps and page speed. Fixes ship directly when I have access, or as tickets your team can merge.
- Weeks 4–8
Make your best pages quotable
Your highest-value pages are restructured around the questions from the baseline, with the clear answers, specifics and schema that get passages cited. New pages fill the gaps where competitors are being recommended.
- Weeks 6–12
Get the rest of the web to agree
Correcting your profiles and entity data, cleaning up stale facts, and earning mentions on the sources the baseline showed models trust in your market.
- Monthly
Re-run the prompts and report
The same prompt set runs every month, alongside your Google rankings, so progress is a before-and-after on the questions that matter — not a vanity score.
Never at the expense of Google
Prompts your buyers use, not ones that flatter
Measured, with the caveats stated
No tricks that get you dropped
Models change, so the work continues
You own everything
What buyers ask before they start
- What is the difference between SEO, AEO and GEO?
- SEO (search engine optimization) earns rankings on results pages like Google. AEO (answer engine optimization) earns a place in direct answers — featured snippets, AI Overviews and voice assistants. GEO (generative engine optimization) earns mentions and citations in answers written by large language models such as ChatGPT, Gemini, Claude and Perplexity. The three overlap heavily, and they work best as one strategy.
- We already rank well on Google. Why are we not showing up in ChatGPT?
- AI answers draw on more than rankings. Models favour sources they can crawl without running JavaScript, passages that answer a question cleanly on their own, and brands described consistently across review sites, directories, forums and publications. A site can rank on Google and still fail every one of those tests.
- Will AI optimization hurt our existing Google rankings?
- It should not, and every change is checked against current rankings before it ships. Most of the work — faster, server-rendered pages, better structured data, clearer answers — also helps traditional SEO. Nothing that already ranks gets rewritten without a reason.
- How do you measure visibility in AI answers?
- With a fixed set of prompts your buyers actually ask, run repeatedly across ChatGPT, Gemini, Claude, Perplexity, Copilot and Google AI Overviews. Each month reports how often you are mentioned, whether you are cited with a link, which competitors appear instead, and how you are described.
- How long until we see results?
- Technical fixes can change what AI search tools retrieve within weeks. Content and off-site authority typically take two to four months to show up consistently, and changes to what a model knows without searching arrive only as new model versions are trained.
- Can you guarantee our brand will be recommended?
- No, and be wary of anyone who does. AI answers vary between runs and change with every model update. What is guaranteed is a measured baseline, the work that moves it, and monthly reporting on the same prompts so you can see exactly what changed.
Find out what AI says about you.
Every engagement starts with an AI visibility audit of your brand against the questions your buyers ask. You keep the findings whether or not you hire me to act on them.
- How often ChatGPT, Gemini, Claude and Perplexity name you today
- Which competitors they recommend instead, and the sources they cite
- The technical and content fixes that would move it, in priority order