Last year a friend asked her phone where to get the best coffee in Wanstead. Not Google. Her phone. She had the ChatGPT app open and asked it out loud, the way you would ask a friend who knows the neighbourhood.
The app gave her three names. I recognised two of them. The third I had never heard of.
All three had one thing in common: well-written, well-structured information about what they did and where they were. Not just a Google Business Profile filled in properly. Something deeper than that.
The café she ended up in that morning was not the one with the best rating on Google. It was the one the algorithm could explain.
The shift
Three years ago, local discovery was straightforward.
You typed something into Google. Google showed you a map. You clicked the profile with the most reviews or the highest stars. That still happens. But it is no longer the whole picture.
More and more people, especially younger customers but not only them, now start with a question. They ask ChatGPT, Perplexity, or Google's own AI Overview. They describe what they want, often in a sentence or two. "Where can I get a proper sit-down lunch near Wanstead station?" "Is there a good independent bookshop in Loughton?"
These tools do not show a list of ten results with paid ads mixed in. They make a recommendation. One, maybe three. With reasons attached.
If you are not in that recommendation, you are invisible in a way you cannot see on your own analytics.
What is actually happening
What the AI assistants are doing when asked for a recommendation.
When someone asks an AI assistant for a local recommendation, the tool is not just pulling from Google. It is drawing on everything it has read: your website, your Google Business Profile, the structured data behind your site, mentions in local articles or directories.
Crucially, it is looking for clear, coherent answers to the questions people ask. Not reviews. Not star ratings. Answers.
"What kind of food do you serve?" Your website should answer that, clearly and naturally. "Are you good for a quiet lunch?" Your Google Business Profile description should address that. "Do you cater for dietary requirements?" You would be surprised how many restaurants do not have this written anywhere in a way a machine can read.
The businesses that the AI recommends are not necessarily the most popular. They are the ones that have given the clearest picture of who they are and what they do.
The good news
This is not a technical problem.
I say this to every business owner I work with: you do not need to understand how large language models work. You need to tell the truth about your business, clearly, in the right places.
A Google Business Profile with a proper description, filled-in services, and recent posts is not just good practice for Google Maps. It is a data source the AI assistants draw from.
A website with a proper About page (one that describes your business in plain English, mentions where you are, what you do, and who you are for) is not just a brochure. It is the document that earns you a recommendation.
Free audit
Want to know how your business appears to AI assistants?The free audit looks at what the AI assistants currently say about you, where the gaps are, and what to fix first. Posted or emailed within a couple of days. No follow-up unless you ask.
Request your free auditWhere to start
Three things you can check yourself this week.
First, read your own Google Business Profile as if you had never heard of yourself. Does it tell a complete stranger what you do, where you are, and what makes you worth visiting? If not, fix it this week.
Second, read your website's About page. Not the home page: the About page. Is it just a paragraph about how long you have been going? Or does it actually describe your business: the food you cook, the products you stock, the clients you take on?
Third, ask ChatGPT what it knows about you. Go to chat.openai.com, type "Tell me about [your business name] in [your area]" and read what comes back. It is not always accurate. But it tells you what the machine has found, and what is missing.
Most businesses have three or four things that are easy to fix and are holding them back. Not difficult technical things. Writing things. Description things. Completeness things.
That is the work.