When buyers ask AI a question, are you in the answer?
Answer Engine Optimisation (AEO) is the practice of making sure AI engines — ChatGPT, Claude, Perplexity, Gemini and Google's AI Overviews — describe, cite and recommend your business when potential customers ask high-intent questions.
How buyer research has changed in 24 months
Ten blue links
Google returned a list of pages. Buyers clicked, compared 3–5 sites, and made a decision. SEO was about ranking page titles and earning backlinks.
The shift to AI answers
ChatGPT crossed 100M users in 2 months. Perplexity, Claude and Gemini followed. Google launched AI Overviews. Buyers stopped clicking — they read the AI summary instead.
The AI answer is the new homepage
Up to 30% of buyer research now starts inside an AI engine. The AI's answer — including which brands it names — replaces the old search results page entirely.
The signal you used to track is gone
In the old world, you measured success by:
SERP rank
Where do I sit on page one of Google?
Click-through rate
What percentage of impressions click my listing?
Backlinks
Who links to me, and how authoritative are they?
Whether ChatGPT mentions you when someone asks "best family hotels in Bali."
Whether Anthropic (Claude) cites your site as a source.
Whether Google (Gemini) named your competitor and not you.
What an AI answer looks like — and why it matters
Below is a stylised example of how a single buyer question now gets answered. The "old" SERP is the blue-link list everyone knows. The "new" answer engine result is what the user actually reads. Notice that the AI's answer has two layers: the brands named in the prose, and the clickable source citations underneath. Both matter, and they're scored separately.
2018 — Old SERP
If Hilton ranks, the family sees their listing. They click and compare.
2026 — AI Overview / ChatGPT
For families visiting Bali, the most popular options are large international resorts like Marriott (multiple Bonvoy properties in Nusa Dua), Hyatt (Regency Bali, well-reviewed for kids' clubs) and The Mulia. Airbnb and Vrbo also list hundreds of family villas in Seminyak and Ubud with pools and kitchens.
Hilton is invisible — not named in the answer text, and not cited as a source the user can click. The family books elsewhere without ever seeing a Hilton listing.
The two layers explained: AI engines return prose and a list of source URLs. The brands named in the prose are who the AI recommends. The cited URLs are what the AI trusts — they're real, clickable links the user can follow. A brand can be named without being cited (often: well-known, but no obvious source page) or cited without being named (often: a comparison page that lists you among options). Strong AEO covers both.
What you actually want to measure
Why it's different from SEO
SEO is about ranking pages
Optimise titles. Earn backlinks. Match keywords. Win the click. The page that ranks #1 wins.
AEO is about being included in the answer
Be cited as a trusted source. Be named in the prose. Be the brand the AI recommends. There is no #1 — there's "named or not named."
What changes if you ignore AEO
- Your brand traffic stays flat or shrinks because AI answers steal the click your homepage used to get.
- Your category traffic goes to whichever competitor the AI happens to name — typically the 2–3 brands AI engines have decided are the "default."
- Your review and comparison traffic gets summarised and dismissed by the AI before the buyer ever clicks a real review.
- Your brand-defence position erodes — if a competitor pays for SEO that ranks an unflattering comparison, the AI may quote it.
Where to start
Run a free 8-question preview to see whether you appear in Google AI Overviews today. It takes 30 seconds and costs nothing. The full audit shows you the same picture across Claude, ChatGPT, Perplexity and Gemini, plus exactly what to publish to fix the gaps.
Frequently asked about AEO
What does AEO stand for?
AEO stands for Answer Engine Optimisation. It's the practice of making sure AI answer engines — ChatGPT, Claude, Perplexity, Gemini and Google AI Overviews — name your brand, cite your domain and recommend you when buyers ask high-intent questions.
Is AEO the same as SEO?
No. SEO optimises for ranking in a list of blue links on a search engine results page. AEO optimises for being named and cited inside the synthesised answer that an AI returns. The fundamentals overlap (fast pages, semantic HTML, canonical URLs) but the success metric is different: in SEO you win clicks, in AEO you win mentions and citations even when nobody clicks.
What metrics matter in AEO?
Four headline metrics. (1) Visibility — the percentage of AI answers that name your brand. (2) Citation rate — the percentage that link to your domain as a source. (3) Competitor share-of-voice — who gets named or cited instead of you. (4) Sentiment + accuracy — whether the AI describes you positively and correctly, with hallucination flags for false claims.
Which AI engines should I track?
The five that drive most current buyer research: Google AI Overviews, ChatGPT, Claude, Perplexity and Gemini. Google AI Overviews now appears on the majority of high-intent commercial queries on mobile; ChatGPT and Perplexity are where buyers go for explicit recommendation questions; Claude and Gemini are growing fastest in B2B.
How do I improve my AEO?
Start by measuring — run an audit to see what each engine says about your brand today. Then prioritise: (a) publish FAQPage and Organization JSON-LD schema so engines can extract factual claims, (b) write Q&A-style pages that answer your buyers' real questions in 2–3 sentences each, (c) make sure robots.txt allows GPTBot, ClaudeBot and PerplexityBot, (d) publish an /llms.txt summary, and (e) build entity signals (sameAs links, ContactPoint) so engines disambiguate your brand from similarly named ones.
How long does AEO take to work?
Faster than SEO. AI engines re-crawl high-signal pages in days, not months, and re-generate answers on every query. Schema fixes (FAQPage, Product, Organization) and content additions typically show up in AI answers within 1–4 weeks. Entity changes (sameAs, ContactPoint) take a bit longer because they have to propagate through the engines' knowledge graphs.