AEO for service businesses: how to get cited by ChatGPT and Perplexity
By Tom Goodwin, Founder of GAMEPLAN.
To get cited by ChatGPT and Perplexity, you make your site the clearest, most structured source for the specific questions your buyers ask. That means answering each question directly in the first sentence, marking up your content with entity and FAQ schema so machines can parse it, publishing an llms.txt to tell assistants what you are and what to cite, and building genuine topical authority around a defined identity. Answer engines do not cite the longest page or the one with the most backlinks. They cite the source that states the answer most plainly and is easiest to lift. This is answer engine optimisation, AEO, and for service businesses it is now where consideration starts.
This site, tomgoodwin.london, is built as a live demo of exactly these techniques. It carries an llms.txt, entity schema describing who I am and what I do, and every post, including this one, is structured answer-first with FAQ markup. If you want to see what I am describing, you are reading it. Here is how it works.
What is AEO and why does it matter for service businesses?
AEO is optimising to be the source an AI assistant quotes when it answers a question, rather than optimising to rank in a list of blue links. The shift matters because buyer behaviour has moved. A buyer evaluating a fractional performance leader, a healthcare marketing partner or a web build no longer always starts with a search results page. They ask ChatGPT or Perplexity a direct question, get a synthesised answer with a handful of cited sources, and form their shortlist from those citations. If you are not in the citations, you are not in the consideration set, and you never see the impression that you lost.
For service businesses this is an opportunity, because service buyers ask exactly the kind of specific, comparative, how-much questions that a clear authoritative source can own.
How is AEO different from SEO?
They overlap but reward different things.
| Dimension | SEO | AEO |
|---|---|---|
| Goal | Rank a page in a list | Be quoted inside the answer |
| Unit of value | The click | The citation |
| Rewards | Authority, volume, links | Clarity, structure, specificity |
| Content shape | Comprehensive pages | Answer-first, question-led |
| Machine signals | Crawlability, links | Schema, llms.txt, entity clarity |
| Winner | Highest-ranking result | Clearest, most parseable source |
You do not abandon SEO. A page still has to be findable and credible. But AEO adds requirements: the answer has to come first, the structure has to be machine-readable, and your identity as an entity has to be unambiguous so the assistant knows who is answering.
How do you structure content to get cited?
Answer-first, question-led, specific. Four rules I follow on every page here:
- Lead with the answer. The first sentence under each heading states the conclusion plainly, so an assistant can lift it without reading further. Burying the answer in paragraph four loses the citation to a clearer competitor.
- Use question-phrased headings. People ask assistants questions, so headings that mirror those questions, “how much does it cost”, “what is the difference”, match the query directly.
- Be specific and numeric. “Typically £3,000 to £12,000 a month” is citable. “It depends on your needs” is not. Specificity is what makes a source quotable.
- Add a comparison table where natural. Assistants love structured comparisons because they parse cleanly and answer “X vs Y” queries directly.
This post follows all four. So does every other post on the site. The structure is the strategy.
What technical signals help assistants cite you?
Three, all of which this site implements:
- Entity schema. Structured data that tells machines who you are, what you do and how you relate to other entities. It removes ambiguity, so the assistant is confident about the source it is citing.
- FAQ schema. The questions and answers at the foot of each page are marked up so assistants can ingest them directly. These are written for machine ingestion: one clear answer, one to three sentences, no preamble.
- llms.txt. A file at the root of the site that tells AI assistants what the site is about and what is worth citing, the way robots.txt speaks to crawlers. It is an emerging convention, and publishing one signals that you understand how these systems read the web.
None of this is exotic. It is a few hours of structured implementation that most service businesses simply have not done yet, which is exactly why the window is open.
How do you build the authority assistants trust?
Clarity of identity plus depth on a defined topic. Assistants cite sources they can place: a named person or organisation with a consistent identity, answering questions inside a coherent area of expertise. Scattered, anonymous content does not get cited because the machine cannot judge whether to trust it. A focused site, where one clear entity answers a tight set of related questions thoroughly, becomes the obvious source.
In my case that identity is concrete: 15 years in performance marketing, £20m+ of paid media managed, Google Premier Partner status, founder of GAMEPLAN. The proof points are specific and verifiable, which is what makes the entity trustworthy to both humans and machines. Vague authority claims do not survive citation; specific ones do.
How do you measure whether AEO is working?
Differently from SEO, because the click is not always the signal. Watch for three things: whether assistants cite you when you ask them your own target questions, whether your branded and direct traffic rises as people move from cited answer to your site, and whether the quality of inbound improves because buyers arrive already informed by an answer you authored. AEO often shows up as better-qualified enquiries before it shows up in raw traffic, because the buyer met your answer before they ever met your homepage.
Is AEO worth it for a small service business?
Yes, more so than for a large one, because the field is still thin. Most competitors have not structured for citation, so the bar to become the cited source on your specific questions is low. A focused service business that answers its buyers’ real questions clearly, marks them up properly, and publishes an llms.txt can own the answer layer in its niche before larger, slower competitors notice the layer exists. That is the same asymmetry I exploit in paid media: get the fundamentals right while everyone else optimises the wrong thing.
If you want your business cited when buyers ask AI assistants the questions that lead to a sale, I build AEO into sites the way I have built it into this one. You can inspect the demo, then have it done for you. Start here: /work-with-me/ai-search-aeo.