AEO / AI Search
AEO: how to get your business cited by ChatGPT, Claude & Gemini
Answer Engine Optimization is how brands get named inside AI answers. Here is how the engines choose their sources — and how to become one.

In this article
- Search didn't die — it moved
- How answer engines actually pick sources
- The on-site work that moves the needle
- How to measure something you can't see in analytics
- The "Ask the Engines" audit, step by step
- The paid pincer: the slot below the answer is now for sale
- Why the window matters
- Frequently asked questions
Search didn't die — it moved
For twenty years the local-marketing game was: rank on Google, earn the click, win the customer. That game is quietly ending. A growing share of buyers now ask ChatGPT, Claude, Gemini, or Perplexity a full question — "who's a reliable plumber near me that won't overcharge?" — and act on the single answer they get back.
If the AI doesn't mention your business in that answer, you were never in the running, and you'll never see it happen in your analytics. Answer Engine Optimization (AEO) is the practice of structuring your content, data, and reputation so AI tools quote you, recommend you, and name you.
How answer engines actually pick sources
Modern AI engines don't store one fixed ranking. When asked a question, they retrieve relevant, trusted content from across the web, then generate an answer and choose which sources to cite — a process technically called retrieval-augmented generation. Get retrieved, and get trusted, and you get named.
Research on citation behavior keeps pointing at the same levers: demonstrated experience and expertise, structural clarity (content written in clean, directly-answerable sections), and factual specificity — verifiable numbers, dates, and named facts get cited while vague claims get skipped.
The on-site work that moves the needle
Most AEO work is unglamorous structure. The good news: it is entirely within your control.
- Answer-first sections: every service page should contain a two-to-three-sentence definitional block an AI can lift verbatim — what you do, for whom, where.
- FAQ with schema: real questions, tight answers, marked up with FAQPage structured data generated from the same content users see.
- Entity clarity: consistent business name, service area, and Organization/LocalBusiness schema so the engines know exactly who you are.
- Open the gates: allow the AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended) in robots.txt, and publish an llms.txt summary of who you are and what you offer.
- Reputation off-site: reviews, local citations, and press are retrieval sources too — engines cross-reference them before recommending anyone.
How to measure something you can't see in analytics
AI referrals do show up (chatgpt.com and perplexity.ai referrers are climbing in most local accounts we see), but the bigger effect is invisible: a customer hears your name in an answer, then searches for you directly or calls. So we track three proxies — branded-search volume, AI-referrer sessions, and a monthly "ask the engines" audit where we run the customer questions that matter and record who gets named.
That last one is the scoreboard, and it deserves its own explanation.
The "Ask the Engines" audit, step by step
Once a month, run the same five questions a real customer would ask, across the four engines that matter — ChatGPT, Claude, Gemini, and Perplexity — and log who gets named. The five questions, templated for any trade: "who is the best {trade} in {city}?", "who should I call for {emergency} in {city} right now?", "{your company} vs {top competitor} — which is better?", "how much does {your core service} cost in {city}?", and "is {your company} legit? what do reviews say?".
Score each cell three ways — named, cited-as-source, or absent — in a plain spreadsheet: twenty cells, ten minutes, dated. The first month is your baseline and it is usually humbling. What makes the audit powerful is the trend: after ninety days of the structural work above, the "absent" cells start flipping, and you can put that sheet in front of anyone who asks what AEO is buying. No tool required, no vendor dashboard — the engines themselves are the scoreboard.
The paid pincer: the slot below the answer is now for sale
Since this article first ran, the game gained a second board: OpenAI now sells sponsored placements inside ChatGPT — clearly labeled ads below the answer, bought through its Ads Manager. The distinction matters: no amount of ad spend puts you inside the answer itself (OpenAI runs ads on separate systems from the model), and no amount of AEO buys you the sponsored slot. They are different real estate on the same screen.
Which is exactly why they compound. The citation earns trust; the ad guarantees presence; a business running both owns the conversation twice while competitors debate whether the channel is real. The organic work in this article is the slower, compounding half — the paid half has its own playbook.
Why the window matters
SEO took a decade to become a knife-fight. AEO is early: most local businesses haven't restructured a single page for it, which means the brands that do the structural work now get cited by default. Early cited sources also tend to keep getting retrieved — visibility compounds.
We optimize for both search and answer engines together, because they feed each other: strong SEO content is the fuel, AEO structure is the delivery system.
Frequently asked questions
What is the difference between AEO and SEO?
SEO earns you a ranked position a human clicks; AEO earns you a mention inside the answer an AI generates. They share most of the underlying work — clear structure, real expertise, consistent entity data — but AEO adds answer-shaped formatting (definitional blocks, FAQs with schema) and open access for AI crawlers. Done right, one effort feeds both.
Which AI engines matter most for a local business?
ChatGPT for sheer usage, Google's AI Overviews because they sit on top of the searches you already care about, and Perplexity because it cites sources aggressively. Claude and Gemini round out the audit set. Run the five-question audit across all of them monthly — the engines disagree with each other, and that disagreement tells you where the gaps are.
How long does it take to get cited by AI engines?
The structural work — answer blocks, FAQ schema, entity cleanup, crawler access — typically starts flipping audit cells within one to three months, with reviews and citations compounding after that. It moves faster than classic SEO because the field is emptier: most competitors have not restructured a single page yet.
Can I pay to appear in AI answers?
Not in the answer itself — the engines do not sell placement inside generated responses, and OpenAI explicitly runs its ads on separate systems from the model. What you can buy is the clearly-labeled sponsored slot below ChatGPT's answers. The mention is earned; the ad is bought; serious operators work both.
How do I measure AEO if AI answers don't show in analytics?
Three proxies: branded-search volume (people hear your name in an answer, then Google you), AI-referrer sessions (chatgpt.com and perplexity.ai show up as referrers), and the monthly ask-the-engines audit — twenty scored cells that turn an invisible channel into a trend line.
Is your expertise easy to find?
Let’s review search visibility, source quality, and the questions your content answers. Get a clear view of what to fix first.
- Your business, your numbers
- Practical next steps
- A conversation with our team
No commitment. Yours to keep.





