Your Business Is Probably Invisible to AI Search. That's a Sales Problem.
We are living in a zero click reality. Most searches on Google now end without a single click to a website. And when generative AI answers the question directly, your business name either appears in that answer – or it does not exist for that customer.
AI optimisation used to mean making your website friendly for search engines. Now it means making your business visible to generative AI: Google SGE, ChatGPT, Perplexity, and Bing AI. Most local businesses have not made the switch. That is a sales problem.
If you search for your own business name plus your city in Google right now, you will likely find your website. That is not the problem.
The problem is what happens when a potential customer does not know your name yet. They type a question like:
- "affordable roof repair north Austin"
- "emergency plumber near me open Sunday"
- "best commercial electrician for a factory shutdown"
On page one of Google, the first thing they see may not be ten blue links. It may be a generative AI answer. A paragraph written by Google SGE, a summary from Perplexity, or a direct answer inside ChatGPT.
If your business is not named in that answer, you are invisible to that customer in that moment.
That is what AI citations are. For local businesses with high-ticket services, being invisible to AI search could mean losing significant revenue – in some cases, five figures per month. That is not a guaranteed figure for every business. But the risk is real.
What an AI Citation Actually Is
A traditional citation is any mention of your business name, address, and phone number on the web. A directory listing. A Yelp page. A Chamber of Commerce member directory.
An AI citation is different. When you ask ChatGPT, Google SGE, Perplexity AI, or Bing AI (Copilot) a question, the machine does not "look up" your business card. It synthesises an answer from training data or live web sources. If your business appears in those sources – and appears credibly – the model may name you.
The old game was about ranking. The new game is about being cited.
Example: User asks: "Who does metal roof repairs in south London?"
AI answer: "Several roofers offer metal repairs. South London Roofing Specialists (often cited on local forums) and MacDonald Roofing (noted for rapid response on Checkatrade) are frequently mentioned. For complex historical properties, some owners use…"
Notice: The AI did not just repeat the #1 Google result. It mixed forum mentions, directory citations, and reputation signals. Your own SEO optimised blog post, if not supported by third party mentions, may be ignored.
That last point is the one most business owners miss.
The Visibility Shift Most Local Owners Miss
I have been doing this work since 1998 – ecommerce SEO, then reputation management, and for the last few years, optimising websites for AI citations. My name is John MacDonald. I run Mac Productions in London.
What I have observed is not a theory. It is a structural change to page one of Google. Generative results are replacing traditional organic listings – zero click search, in industry terms. So are Google Shopping carousels, Maps packs, and "People also ask" boxes.
At Google I/O, the company announced its intention to populate the whole of page one with AI-generated answers. The traditional blue links to websites? I assume they drop to page two. Not pushed below the fold. Not scrolled past. Page two.
What is Generative Engine Optimisation (GEO)?
You may have heard the term Generative Engine Optimisation, or GEO. It is the emerging practice of optimising content specifically for AI generated answers, rather than traditional blue link rankings. AI citations are a core part of GEO. The three factors below are where I start with every client.
Why High Ticket Businesses Feel This First
A coffee shop survives on foot traffic and habit. A roofer, a commercial electrician, a luxury home builder, a dental implant clinic – these businesses cannot. Their customers research first. They ask multiple questions. And increasingly, they start with AI.
To illustrate: if your average job is worth $10k and you lose even a couple of jobs per month to competitors who appear in AI answers, that adds up fast. This is an illustration, not a guarantee. But it is why high ticket business owners are paying attention.
I will not pretend every business loses money this way. Some niches are less affected. Some geographic areas have lower AI adoption. The evidence is mixed and context dependent. What I can say, from running many audits for local businesses over several years, is that the pattern is clear: businesses with strong third party citation profiles appear in AI answers more often than those without.
Three Critical Factors for AI Optimisation Readiness
I built software to measure how well a website is optimised for both traditional SEO and AI visibility. Why? Because I needed to work on a site with over 5,000 pages. A checklist would have been useless. I needed something that could crawl, measure, and verify.
Anyone can claim "I will fix your AI citations." But how do you verify the work? You need citeability – a measurable quality of how well your website feeds the machines.
Here are the three on site factors my tool measures. If your website gets these right, you have a foundation. If not, nothing else matters.
1. Structured Data Completeness
Schema markup – JSON-LD, specifically – tells AI models exactly what your business is, where it serves, what services you offer, and your hours. It is the difference between an AI guessing and an AI knowing.
If your site has incomplete or incorrect schema, you are asking AI to guess. AI does not guess well.
2. FAQ Content for Long Tail Question Matching
AI models love questions and answers. Not just any answers – clear, direct, factual answers that a model can quote verbatim.
When you publish well structured FAQ content, you are feeding the synthetic brain with long tail keyword questions and their matching answers.
3. An LLM.txt File
You have probably heard of robots.txt. That tells search engines what they can and cannot crawl. LLM.txt is newer. It tells large language models what on your site is worth reading.
Most websites do not have one. That is a problem. My software checks for it, validates it, and flags what is missing.
What About Off Page Factors?
I should be clear. AI citations do not come only from your website. There are off page data types that AI counts too. My software does not measure the open web. It measures your website. But you still need to know what else matters.
Here are two off page factors I check manually for every client.
A. NAP Consistency Across the Long Tail
Everyone knows to keep name, address, and phone consistent on major directories. The killer is the long tail: old forum profiles, archived directory listings, supplier pages, community event listings. One inconsistent address – even "Suite 200" versus "Ste 200" – can fragment your citation footprint.
B. Third Party Mention Density (Not Just Links)
AI models care about mentions, not just backlinks. A forum post that names your business but does not link to it is still a citation. A news article that spells your address correctly but uses a no follow link is still valuable. You have to feed the synthetic brain with real, third party signals. Most SEO tools ignore this entirely.
Your 30 Second Free AI Visibility Test
You do not need to guess whether your business is invisible to AI. I built a free AI visibility test. It takes about thirty seconds. It checks your business against the same factors my full audit measures – third party mentions, structured data, NAP consistency – and gives you your AI Readiness Score.
No credit card. No call from a salesperson unless you ask.
You will see:
- What is missing
- What is correct
- A clear path from where you are to where you need to be
Because I operate from London under UK GDPR and my own privacy and cookie policies, your data is not sold or used to train anything.
What to Do Next
If your score is low, do not panic. AI citations are fixable. They are not mysterious. They require systematic work: cleaning up old mentions, adding structured data, and earning (or reactivating) genuine third party references.
If your score is high, maintain it. Rerun the report every quarter. AI models change. Citation sources change. What worked six months ago may not work today.
The single worst action is to do nothing while assuming "my SEO is fine." Your SEO might be fine for 2019. For generative search? Fine is not enough.
