AI is the new word of mouth for SMBs. Here's how to win it.

Picture the moment a customer needs you.

Their boiler is leaking. A pipe burst under the sink. They need a dentist who can see their kid today, or an accountant who actually understands restaurants. For decades, that moment had two outcomes: they asked a neighbour, or they Googled it and scrolled through a list.

Now there's a third, and it's taking over fast. They open ChatGPT, Gemini or Claude and type the whole situation: "My boiler is leaking, who can come today in the 11th arrondissement?" And the AI doesn't hand back ten links to compare. It recommends one business.

For a small local business, this is the biggest change in how you get found since Google existed. Because that AI has quietly become the most powerful word of mouth in the world. It never sleeps, it answers thousands of people a day, and when someone asks it to recommend a business like yours, it gives a name.

The only question that matters now is whether that name is yours.

The good news: word of mouth was never about budget

Here's why small businesses should be excited rather than scared.

Word of mouth never rewarded the company with the biggest ad spend. It rewarded the one people could describe precisely. Nobody ever said "you should call that trusted heating solutions provider." They said "call Marc, he fixed my boiler in an hour last winter." Specific, concrete, memorable.

AI recommendations work the same way, and that is a structural advantage for small business. A national chain can buy awareness. It cannot buy specificity. It talks about "trusted heating solutions." You talk about repairing a Vaillant boiler within two hours in central Paris. For years that narrowness felt like a limitation, the small operator who only does one thing, in one place, for one kind of customer. In the AI era it becomes your sharpest weapon, because when someone asks a precise question, the AI isn't hunting for the biggest brand. It's hunting for the best-matched one.

Specificity used to be a ceiling. Now it's the signal that gets you named.

The catch: this "friend" only knows what it can read

There's one important difference between AI and the neighbour who recommends you.

Your neighbour knows your reputation from living next to it. The AI doesn't. It has never walked past your shop or heard a friend rave about you. It only knows what is written down in the places it can read: your website, the directories you're listed in, the reviews customers leave, the articles or local forums that mention you.

So a brilliant local business with a spotless reputation can still be invisible to AI, simply because that reputation lives in people's heads and not in text a machine can find. The reputation is real. It just isn't legible.

That's the whole game for small business now: turning the reputation you already have into evidence an AI can actually read.

What AI actually reads about your business

When an AI decides which local business to recommend, it isn't flipping a coin. It reads a handful of signals, and they don't all carry the same weight. Most owners obsess over the wrong one.

Your star rating matters less than you think. A row of five stars is just a number, and half your competitors have it too. What the model really leans on is closer to evidence:

1. What your reviews actually say. Not the score, the words. "Great service" tells the AI nothing. "Fixed a leaking radiator in an hour, same day" tells it exactly which question you're the answer to.
2. Whether your details are consistent everywhere. Same name, same address, same phone, same opening hours, across your site, Google, directories, everywhere. When those disagree, the model hesitates, and a model that hesitates recommends someone else.
3. Whether your site clearly says what you do, where, and for whom. Not clever taglines. Plain, findable facts a machine can parse without guessing.
4. Whether a credible source has mentioned you. A local directory, a press mention, a supplier's partner page, a forum thread. Outside confirmation tells the AI you're real and relevant.

None of these is pay-to-win. There is no ad budget here, no agency on retainer. These are the cheapest levers a small business fully controls, and they are exactly where the recommendation is decided.

Your reviews are the new word of mouth, if you ask the right question

Of all those levers, reviews are the one most small businesses are sitting on without using properly.

Most owners chase a rating. Five stars, feel good, move on. But two businesses with identical ratings are no longer equal in the eyes of an AI. The one whose reviews describe real, specific jobs gets recommended. The one with vague praise disappears.

And this is free leverage, because you don't buy it, you ask for it. The change is one sentence. Instead of "please leave us a review," try "tell us what we fixed, and how fast." A review that says "replaced our boiler the same day we called, in August, no fuss" is worth ten that say "very professional." One is a compliment. The other is evidence a model can quote when someone asks for exactly that.

Why this matters far beyond your shop

Step back and the stakes are huge, because small business isn't a niche. SMBs are the backbone of the economy, and arguably its heart. Across France, the UK, Germany and Dubai the story is the same: they are roughly 99% of all companies, most of the jobs, an enormous share of GDP. According to BrightLocal, 84% of consumers use online search to find local businesses, and AI is rapidly becoming part of that journey.

For a global brand, AI search is becoming a new acquisition channel. For a small or local business, it could soon become the primary one. The neighbourhood is going digital again, and this time the person everyone asks for a recommendation is a machine.

What you can do this month

You don't need a budget or an agency to start. You need your reputation written down where a machine can read it.

Make sure your business details are identical everywhere they appear. Make sure your website says, in plain words, what you do, where, and who for. Ask your next ten happy customers not for "a review" but for a specific one: what did you fix, and how fast. And get mentioned somewhere credible in your area, a directory, a local paper, a partner.

The winners in AI search won't necessarily be the biggest brands. They'll be the businesses with the clearest, most specific evidence for the exact question a customer is asking. For once, that's a game a small business can win outright.

So here's the question worth sitting with. If someone asks an AI the exact question your business was built to answer, will it say your name?

That is precisely what AYAN helps you measure and improve: what AI actually says about your business, where it gets you wrong, and which lever to pull first.