Reviews and AI Search: How ChatGPT and Gemini Pick Local Businesses
45% of consumers now use AI tools to find local businesses, up from 6% a year earlier. What AI assistants read when they recommend you, why reviews matter, and what you can do about it.

On this page
- How many people use AI search to find local businesses?
- Where do AI assistants get their information?
- Why detail beats keywords
- What I'd actually do to get recommended more often
- Does review recency matter to AI?
- AI review summaries
- Can you fake your way into AI recommendations?
- A monthly check you can do in ten minutes
Key takeaways
- 45% of US consumers used ChatGPT or other AI tools for local recommendations in 2026, up from 6% (BrightLocal).
- 82% read AI review summaries and 23% would rely on them alone.
- Assistants read review text, so ask customers for detail, never for keywords.
- Matching business details everywhere and a plainly written website are the levers you control.
- Fake reviews, AI-written ones included, break platform rules and, in the US, the FTC rule.
"Does ChatGPT read my Google reviews?" I get some version of that question from clients all the time now.
Short answer: AI assistants like ChatGPT, Gemini and Google's AI Overviews recommend local businesses by pulling together what they can find about them, and customer reviews are one of the richest things they can find. So reviews matter in AI search, probably more than they did in the old list of blue links, because these tools read the words, not just the stars.
What I can't tell you is exactly how each system weighs its sources. Nobody outside those companies can. If someone promises you guaranteed AI rankings, they're selling something. But there's enough to act on.
How many people use AI search to find local businesses?#
More than I expected, and the jump was fast. BrightLocal's Local Consumer Review Survey 2026, a panel of 1,002 US adults, found:
| Finding (2026) | Share of consumers |
|---|---|
| Used ChatGPT or other AI tools for local recommendations | 45% (6% the year before) |
| Trust AI platforms to provide business recommendations | 40% |
| Trust AI recommendations as much as traditional reviews | 42% |
| Read AI-generated review summaries | 82% |
| Would rely only on an AI review summary to decide | 23% |
| Used Google to read local business reviews | 71% (83% the year before) |

Google is still the biggest single source by a distance. It just isn't the only one that matters anymore. And even inside Google, AI is often the first thing people read now: summaries above the individual reviews, and AI Overviews above the search results.
Where do AI assistants get their information?#
Each one works differently, and most don't publish the details. Broadly, they draw on some mix of:
- business listings such as Google Business Profile, Apple Maps and Bing Places
- review platforms, from Google and Facebook to the sites specific to your trade
- your own website
- mentions elsewhere: local news, "best of" lists, directories, forums
- live search results, for assistants that browse while they answer
Reviews stand out in that list because they're the main place independent people describe, in their own words, what you're actually like. Picture someone asking an assistant, "Which dentist near Koramangala is good with nervous patients?" The assistant can only connect your clinic to nervous patients if somebody has written about it.
Why detail beats keywords#
Traditional local search leaned heavily on your rating and review count. AI search reads the text.
A review along the lines of "the dentist explained every step and I wasn't anxious at all" gives an assistant something specific to repeat. A review that just says "Good" gives it nothing.
That changes what you should ask customers for. Not keywords. Asking people to mention "best emergency plumber in Pune" makes reviews sound scripted, and scripted reviews look fake to readers and run against platform rules. Ask for detail instead, with open questions:
- "What did we help you with today?"
- "What would you tell a friend who's thinking of coming to us?"
My review request templates include versions written to draw out that kind of detail.
What I'd actually do to get recommended more often#
There's no switch to flip. These are the things that give AI systems clearer, richer and more consistent material about you, and every one of them also helps with human customers:
- Keep a steady flow of detailed reviews coming in on Google, plus one or two platforms that matter in your industry. See how to get more Google reviews.
- Make your business details identical everywhere. Name, address, phone, hours and categories should match across your Business Profile, website and directories. Conflicting details create doubt for machines as much as for people.
- Write your website plainly. Say what you do, where, for whom and roughly what it costs. Put the answer to each common question in the first sentence of its section, so a system can quote you accurately.
- Add organization details in structured data (name, logo, address, contact) so machines can identify your business. Skip self-serving review markup; review schema and star ratings explains why.
- Reply to reviews properly. A reply like "we've since added weekend appointments" is new information an assistant might pick up.
- Earn mentions on local sites people trust: local press, your chamber of commerce, community groups, trade associations.
And show your reviews on your own website. People who arrive from an AI recommendation still check you out, and seeing the same praise on your site confirms what they were told. Where to put reviews on your website has a placement plan.
Does review recency matter to AI?#
It matters to people, and I think it's a fair assumption that it matters to systems trying to give current answers. BrightLocal found 74% of consumers only care about reviews from the last three months. A summary built on reviews from two years ago describes a business that may not exist in that form anymore. Keep new ones coming; there's more on this in review recency.
AI review summaries#
Google, Amazon and others now put an AI-written summary above individual reviews, and in BrightLocal's survey 82% of consumers said they read them. You can't edit those summaries. You can only shape what goes into them. If most recent reviews mention quick service and friendly staff, that's what the summary tends to say. If one complaint keeps coming up, fixing the actual problem is the only lever that works.
Can you fake your way into AI recommendations?#
No, and trying is risky. Assistants lean on platforms that actively remove fake reviews, and fake reviews break Google's policies everywhere. In the US, the FTC's rule on fake reviews, in force since October 2024, covers reviews generated by AI as well. Even if fake praise got you recommended, you'd be sending yourself customers who expect something you didn't deliver, and they'd say so. Fake reviews and the law covers the rules.
A monthly check you can do in ten minutes#
You can't see inside these systems, but you can ask them the way your customers do. Once a month, try a few prompts in the assistants your customers are likely to use:
- "Recommend a [your service] in [your area]."
- "Is [your business name] in [your area] any good?"
- "What do customers say about [your business name]?"
Note if you're mentioned, what's said, and which sources are cited when the assistant shows them. If an answer is wrong or stale, the fix is almost always upstream:
| What you find | Likely cause | What to do |
|---|---|---|
| You're not mentioned at all | Thin or inconsistent information about you | Complete your profile, build reviews, earn local mentions |
| Old hours or services | Outdated listings or pages | Update every listing and your website |
| Only vague praise | Short, generic reviews | Ask open questions that invite detail |
| A recurring complaint | A real pattern in your reviews | Fix the problem, then reply explaining the fix |
AI search doesn't replace the basics. It raises the stakes on them. Detailed, recent, genuine reviews, consistent business details and a clear website are what give assistants something accurate and good to say about you. And when those recommendations send people to your site, ReviewPine shows them the same real Google reviews there, right where they decide.
Frequently asked questions
How does ChatGPT decide which local businesses to recommend?
Do Google reviews affect AI search results?
Should I ask customers to use certain keywords in their reviews?
Is AI replacing Google for finding local businesses?
How can I see what AI assistants say about my business?
Turn your Google reviews into your best salesperson
Show real Google reviews on any page of your website, with a "Review us on Google" button that brings in new ones. On paid plans, new reviews appear by themselves. Set up in about five minutes.
- Works on WordPress, Shopify, Wix and any site
- Paid plans update by themselves
- No cookies, loads after your page


