Does AI recommend your business?
More and more customers ask ChatGPT, Claude, and Gemini for recommendations instead of searching. We'll ask an AI the way a customer would — and show you what it says.
Why customers now ask an AI instead of searching
“Find me a good physio in Newtown that's open Saturday.” A few years ago that was a search box, ten blue links, and a customer who did the sorting themselves. Increasingly it is a question put to ChatGPT, Gemini, Claude or Perplexity — and the answer comes back as a short list of two or three named businesses. There is no page two. If you are not on that list, you were not considered.
This matters most for exactly the businesses least likely to be watching it: local, owner-operated, competing inside a suburb rather than a category. The shortlist is short, so the difference between being on it and not is the entire enquiry.
How an AI decides which businesses to name
An assistant answering that question is not consulting a directory it maintains. It is drawing on what it has read about your area — your website, your listings, reviews, directory entries, news, forum threads — and on live search results it fetches while answering. Both routes reward the same thing: a business that is described clearly, in text, in more than one place.
That is why a business with a real website naming its services, prices, suburb and opening hours tends to get mentioned, while one that exists only as a social profile often does not. Photos of your work are invisible to a language model. The sentence “we do balayage in Bondi and open at 8am Saturdays” is not.
Two practical consequences. First, thin presence is a real handicap: if there is nothing readable about you, there is nothing to cite. Second, being invisible in a quiet area is a temporary state, not a permanent one — the businesses that publish properly first are the ones that get named as assistants catch up.
What to do about it
Publish the facts a customer would ask for, as text, somewhere you control: what you do, what it costs, where you are, when you are open, and what makes you the right choice. Keep your listings consistent with it — the same name, address and phone number everywhere, because contradictions make a business look less reliable to a model deciding whether to name it.
Then make sure the enquiry survives arrival. Being recommended is worth nothing if the call rings out — and the customer who was told about you will simply ask the assistant for the next name on the list. That is the part an AI receptionist covers: the phone answered, the price quoted from your real list, the booking taken.
Common questions
Does this check every AI assistant?
No. It asks one assistant the way a customer would and shows you the answer. Different assistants draw on different sources, so treat the result as a sample rather than a verdict — if you are missing from one, you are usually missing from most, for the same reason.
My business is on Instagram and Facebook. Is that enough?
Usually not on its own. Social profiles are largely images and short captions, often behind a login, and much of what makes you findable — services, prices, hours, suburb — never appears as readable text. A simple website carrying those facts does more for AI visibility than a large following.
How long does it take to start being mentioned?
Weeks rather than days, and it depends on how much already exists about you. Assistants that fetch live search results can pick up a new site quickly; the knowledge baked into a model updates far more slowly. Publishing the facts clearly is what shortens both.
The check said AI knows nothing about my area. Is that bad?
It is an opportunity more than a problem. It means no competitor has established a presence an assistant trusts either, so the first business in the area to publish properly tends to become the one that gets named.
What does MAVRK do about this?
MAVRK builds your business a real website with your services, prices and booking — the readable presence assistants learn from — and then answers the calls and WhatsApp messages that come from it, so a recommendation actually turns into a booking.