How AI tools pick which small business to recommend

You ask ChatGPT for a good Italian restaurant in Brighton and it names three. You ask Claude for a recommended physio nearby and it picks two. The question that nags is the obvious one: why those? Why not yours, or the one round the corner that everyone in town swears by?

The honest answer is that AI tools don't pick favourites. They piece together a recommendation from whatever they can find about each business — and most websites give them very little to work with. By the end of this post, you'll know the three signals AI tools weigh most heavily, why bigger businesses don't always win, and what to do if your business isn't sending those signals clearly today.

What "recommend" actually means inside an AI tool

When someone asks ChatGPT, Claude, Gemini or Perplexity for a recommendation, the tool isn't pulling from a curated list. It's reading public information in real time — your website, your business listings, news mentions, third-party write-ups, social posts — and deciding which businesses match the question best.

It's not a popularity contest

Bigger brand, more customers, more years in business — none of that automatically wins. AI tools care about whether they can confidently describe what you do, where you do it, and why someone might choose you over the next option. A driving instructor in Cardiff who explains exactly which test centres they cover and what their first-time pass rate is can beat a national chain that just lists "driving lessons across the UK".

It's an evidence test

Think of it like a friend giving a recommendation. They won't suggest you if they can't remember what you actually offer, or whether you're any good. AI tools work the same way — they need clear, recent, specific evidence before they'll put your name forward. AI My Site shows you exactly what evidence Claude, ChatGPT and Perplexity can find about your business right now, and what's missing on the page.

The three signals AI tools weigh most heavily

Across thousands of recommendation queries, the same three signals come up again and again. Get these right and you stop being invisible.

Specific, plain-language descriptions of what you do

AI tools want a direct answer to "what does this business do, and for whom?" Vague hero copy like "your trusted partner for excellence" tells them nothing. A baker in Sheffield who describes themselves as "a small-batch sourdough bakery on Ecclesall Road, supplying eight independent cafés across South Yorkshire" gives the AI everything it needs to bring them up when someone asks for "a sourdough bakery near Sheffield". The more specific the description, the more confidently the AI will quote it.

Recently updated, dated information

AI tools quietly downgrade businesses that look stale. If your last blog post is from 2022, your hours haven't been updated since the pandemic, or your prices read "from £25" with no context, AI tools assume the information might be out of date and reach for a more recent option instead. Even one fresh page a month is enough to signal that you're still trading and your details still apply.

Third-party mentions that confirm what your site says

This is the one most small businesses miss. AI tools don't just read your website — they cross-reference what you say about yourself against what other sites say about you. Local press, directory listings, customer reviews on independent platforms, podcast mentions, partner pages — all of these act as confirmation. A physio in Brighton who's quoted in two local news pieces and listed on a regional health directory will outrank a competitor with a slicker website but no third-party footprint.

Why bigger businesses don't always win

If you've ever wondered how a one-person operation beats a national chain in an AI recommendation, this is why.

Big brands are often less specific

National chains tend to write for everyone, which means they're specific to no one. When their site says "we serve customers across the UK", an AI tool can't confidently match them to "kitchen fitter near Norwich". A local kitchen fitter who names the Norwich postcodes they cover and lists three recent jobs by area gets matched first, every time.

Smaller businesses can move faster

Updating your homepage, adding an FAQ, getting a quote into a local paper — these are weeks of work for a marketing team at a chain and an afternoon for you. The businesses winning AI recommendations right now are mostly small, focused operations that have made a few targeted changes. You can check your own progress over time by tracking your AI readiness score against the same questions every few weeks and watching the answers shift.

Common mistakes to avoid

• Writing for search engines instead of people: stuffing the same keywords into every page makes you sound generic to AI tools, not relevant. Write the way a knowledgeable owner would describe their business to a friend.

• Hiding your specifics behind a contact form: prices, service areas, opening hours and qualifications all need to be on your site in plain text. If an AI has to guess, it'll usually pick someone whose details are clearer.

• Ignoring third-party mentions: a business with two paragraphs in a local paper and a mention on a community site will beat a similar business with a prettier website and no outside coverage.

• Treating your homepage as the only page that matters: AI tools often quote from FAQs, service pages and About pages. A thin homepage with everything else buried two clicks deep limits what AI tools can find.

How long this takes to show up

Once you've made the changes, expect to see shifts in how AI tools describe your business within four to eight weeks. The cleanup itself — rewriting your homepage paragraph by paragraph, adding three or four FAQ entries, updating your About page with current details — usually takes a weekend. Earning new third-party mentions is slower; plan for a steady drip of coverage over three months rather than a single sudden change. The businesses that re-scan monthly and follow each new task one at a time tend to see the cleanest results, because they catch small drift before it builds up.

Recommendations from AI tools aren't random and they aren't rigged. They're the result of clear, fresh, well-corroborated information that any small business can choose to provide.

Want to know exactly what else is holding your website back? Sign up at AI My Site and get your complete SEO and AI readiness action plan in minutes — with step-by-step guides written specifically for your website platform.