How to Get Your Business Cited in ChatGPT, Perplexity and Google AI Overviews
Your customers have started asking AI instead of Google. Generative engine optimisation is how you make sure your business is the answer it gives them – and most Glasgow companies have not noticed the shift yet.
Generative Engine Optimisation (GEO) is the practice of structuring your content so that AI tools like ChatGPT, Perplexity, Google AI Overviews and Microsoft Copilot cite your business as a source in the answers they generate. It is not the same as traditional SEO. SEO gets you ranked in a list of blue links. GEO gets you named inside the answer itself, before anyone clicks anything.
That distinction is about to matter enormously, and right now it is a quiet advantage rather than a crowded one.
Why this matters now, not later
The way people search has genuinely changed. When someone asks ChatGPT “who is the best marketing agency in Glasgow” or asks Perplexity “which windows company installs bifold doors in Stirling”, they no longer get ten links to work through. They get a short, confident answer that names a handful of businesses. If yours is one of them, you are on the shortlist before a single sales conversation happens. If it is not, you are invisible to that buyer entirely.
The numbers behind this are hard to ignore. Google AI Overviews now appear in a large share of searches, and when they do, the top organic result loses roughly a third of its clicks. Meanwhile, AI-referred visitors convert far better than ordinary search traffic – several independent studies put the difference at five to nine times higher. The reason is simple: someone who arrives from an AI answer has already done their research inside the tool, so they land on your site with real intent rather than idle curiosity.
Here is the part most businesses miss. The overlap between who ranks on Google and who gets cited by AI is smaller than you would expect. One analysis of 680 million citations found that only 11% of domains are cited by both ChatGPT and Perplexity. Ranking well on Google no longer guarantees you appear in the AI answer. That is a problem if you have ignored GEO, and an opportunity if you move before your competitors do.
How AI engines actually choose who to cite
The mental model that helps most is this: an AI engine does not read your page as a document. It reads it as a list of passages, scores each one independently, and picks the passages that best answer the question in front of it. Your job is to make your passages the easiest and most trustworthy ones to lift.
Three things drive that decision.
Extractability. The engine wants a complete, self-contained answer near the top of a clearly labelled section. A passage that answers a question in its first forty to sixty words, without making the reader scroll, scores far higher than one that builds slowly to a conclusion. Notice how the section under each heading in this article opens with a direct answer – that is not a coincidence, it is the technique.
Structure the machine can read. Question-style headings that match how people actually phrase queries. Schema.org structured data in the page code that tells the engine what your business is and what it does. Clean HTML that a crawler can parse without running JavaScript, because most AI crawlers do not execute it. If your content only appears after a script runs, the engine simply cannot see it.
Consensus and trust. This is the one businesses underestimate. AI engines look for agreement across several independent sources before they cite you with confidence. If your positioning shows up consistently on your own site, in industry publications, on LinkedIn, in genuine community discussions and on review platforms, the engine trusts it. If you exist only on your own website, making claims nothing else backs up, the engine treats those claims with scepticism and recommends a competitor who has built a broader footprint.
The tactics that actually earn citations
None of this is magic. The work is mostly structural, and every one of these is something you can apply to the pages you already publish.
Open every section with the answer. Lead with a self-contained response of roughly forty to sixty words, then expand. Do not save your conclusion for the end.
Phrase your headings as real questions. “How much does a new website cost in Glasgow” beats “Our pricing”. The heading that matches the user’s query gets the passage selected.
Use verifiable statistics and name your sources. The Princeton GEO study found that adding statistics and adding source citations each improved AI visibility by thirty to forty per cent. “This tool is great” gets ignored. “According to a 2026 analysis, AI referral traffic converts five times better than organic” gets cited.
Add FAQ and structured data. FAQ schema makes a page significantly more likely to appear in Google AI Overviews, and models love a clean question-and-answer format. Structured data confirms to the engine exactly what your page is about.
Let the AI crawlers in. Check your robots.txt allows GPTBot and OAI-SearchBot (OpenAI), PerplexityBot, ClaudeBot (Anthropic) and Google-Extended (Gemini). A page that is not fetched cannot be indexed, and a page that is not indexed cannot be cited. This is the precondition everything else depends on, and it is often the thing quietly blocking a business without anyone realising.
Keep it current. Perplexity in particular favours fresh content, citing pages published in the last month at a much higher rate, and current-year signals in titles and headings measurably improve citation rates. GEO is not set-and-forget. It rewards a business that keeps its cornerstone content updated.
Do not treat AI search as one thing
A tactic that works on Perplexity may do nothing on ChatGPT. The engines use genuinely different logic. Perplexity performs a live web search for every query, has no knowledge cutoff, and can cite a well-structured article within hours of publication – which makes it the fastest place to test whether your GEO is working. ChatGPT leans on Bing’s index and established authority, so it moves slower and rewards patience. Google AI Overviews inherit Google’s own indexing pace and lag the longest, often four to eight weeks.
The practical upshot: start with Perplexity for the quick feedback loop, then build the authority signals that ChatGPT and Google reward over the following months. It compounds.
How to know if your generative engine optimisation is working
You cannot improve what you do not measure. Pick three or four questions you most want your business to be the answer to. Once a week, ask them of ChatGPT, Perplexity and Google’s AI, and record whether you were named, how you were described, and who appeared instead. It takes five minutes and it tells you more than any dashboard. Over a few weeks you will see the pattern shift as your content starts getting picked up – Perplexity first, then the others.
Where Pink Pine comes in
We treat SEO and generative engine optimisation as one job now, not two. Rank on Google, and get quoted by the AI. We build the structured data, shape the content around the questions your customers actually ask, sort the crawler access, and put in the consensus signals that make an engine confident your business is the credible answer. We already do this for our clients – one query for a client of ours currently ranks first on Microsoft Copilot – and because most businesses in Scotland have not started, the ground is genuinely still open.
The companies that move on this now will spend the next few years being the name the AI recommends. The ones that wait will be explaining to their AI-using customers why they never show up. It is a rare thing in marketing: a real head start, available to whoever claims it first.
If you want your business to be the one the AI names, start a project with us – or read how we built an AI growth engine for an energy consultancy to see the approach in practice.