Key Points
- BrightLocal’s 2026 Local Consumer Review Survey found 45% of U.S. consumers used AI tools like ChatGPT for local business recommendations, up from 6% a year earlier.
- SOCi’s 2026 Local Visibility Index, covering more than 350,000 business locations, found ChatGPT recommends just 1.2% of local businesses.
- An Omniscient Digital analysis of 23,000+ citations found roughly 77% of the sources AI answers cite about a brand come from other websites, not the brand’s own.
In This Article
- Why does ChatGPT get details about your business wrong?
- Where do AI assistants get their information about local businesses?
- How can you check what AI says about your business?
- What will change for AI answers about local businesses over the next 12 to 24 months?
- What should you fix first?
- What else do business owners ask about AI answers?
Quick Answer
ChatGPT gets businesses wrong because it builds its answer from whatever the web says about you, and most of that lives on other people’s websites. Outdated directory listings, thin website copy and missing reviews all flow straight into the answer. The fix starts with checking what AI actually says.
The shift happened fast. According to BrightLocal’s 2026 Local Consumer Review Survey, 45% of U.S. consumers now use ChatGPT or another AI tool to find local businesses, up from 6% a year earlier. Google’s share fell from 83% to 71% over the same period.
Most businesses haven’t caught up. SOCi’s 2026 Local Visibility Index found ChatGPT recommends just 1.2% of local businesses. In my view, that gap is the biggest blind spot in local marketing right now: customers are asking a new question in a new place, and most owners have never looked at the answer.
Below, I explain where AI gets its facts, why they go stale, and how to check your own business. The habit I’d build is the three-prompt test: ask an AI assistant what your business does, who’s best at what you do in your city, and when you’re open. Then compare every answer to the truth.
Top questions this article answers
- Why does ChatGPT say the wrong thing about my business?
- Where does ChatGPT get information about local businesses?
- How do I check what AI tools say about my business?
Why does ChatGPT get details about your business wrong?
Because it answers from a patchwork of sources, and any stale or vague piece shows up in the reply. AI doesn’t call your front desk to confirm your hours.
A person reading your homepage can scroll, click around and fill in the blanks when a sentence is vague. An AI tool can’t. If your site says “we serve the greater area” instead of naming towns, or lists services without prices or details, the assistant fills the gap from somewhere else.
A common misconception is that AI reads your website first and trusts it most. The reality is different. According to a 2026 roundup of AI search statistics, an Omniscient Digital analysis of more than 23,000 citations found about 77% of the sources cited in AI answers about a brand are off-page: directories, reviews, forums and video. If an old directory lists the hours you had in 2022, that’s the version an assistant may repeat.
Freshness matters too. The same roundup notes that ChatGPT showed the strongest preference for recent content among the engines measured. A business that hasn’t updated anything in two years looks, to an AI, like a business that may not exist anymore.
In practice, wrong answers usually trace back to wrong or thin information somewhere on the web. The takeaway: the fix is rarely one page. It’s the whole trail of information about you.
Where do AI assistants get their information about local businesses?
From search indexes, review sites, directories and your own website, weighted differently by each assistant. There’s no single database you can update once.
Each tool leans on different sources. HubSpot’s guide to ChatGPT and SEO points out that ChatGPT depends on Bing, and recommends setting up Bing Places for Business alongside your Google Business Profile. Many owners have polished their Google listing and never touched Bing.
Here is how the main sources usually feed an AI answer:
| Source | What AI takes from it | Common failure |
|---|---|---|
| Your website | Services, locations, prices, story | Vague copy that can’t be quoted |
| Search listings (Google, Bing) | Hours, address, phone, categories | Bing listing never claimed |
| Review sites | Reputation, recent experiences | Few or old reviews |
| Directories and forums | Confirmation from third parties | Outdated or conflicting details |
Customers also check the work. In Bill Hartzer’s breakdown of the BrightLocal data, 88% of consumers verify what AI tells them. ChatGPT leads at 31% of consumers using AI for local searches, with Google AI Mode at 23%. So a wrong AI answer is followed by a second check, and if your listings disagree with each other, you lose trust twice.
What this means: consistency across every source matters more than perfection on any one of them.
How can you check what AI says about your business?
Run the three-prompt test in two or three assistants, write down every answer and compare it line by line with your real details.
Here’s how I’d run the three-prompt test:
- Prompt 1: “What does [business name] in [city] do?” This checks whether the assistant knows you exist and describes your services correctly.
- Prompt 2: “Who is the best [your service] in [city]?” This checks whether you appear in recommendations, and who appears instead.
- Prompt 3: “What are [business name]’s hours and prices?” This checks the details most likely to be stale.
Do it in ChatGPT and at least one other assistant, such as Gemini or Google’s AI Mode, because each one draws from different sources. Note which facts are wrong and where the wrong version might have come from, such as an old directory or a review that mentions a closed location.
The answers are only a snapshot. To understand why one business gets named and another doesn’t, it helps to see how AI visibility scores are calculated: whether your content answers questions directly, whether it’s structured so an AI can quote it, and whether other sites back it up.
In practice, 20 minutes of prompting will tell you more than a month of guessing. The takeaway: check before you fix.
What will change for AI answers about local businesses over the next 12 to 24 months?
AI answers will become the first thing many customers see, and freshness will matter more than technical tricks. I expect the owners who keep their information current to pull ahead quietly.
| Prediction | Weak signal | Why it matters | Source |
|---|---|---|---|
| AI answers sit above most local results. | AI Overviews now appear on 68% of local business-type queries, compared with 39% for the traditional map pack. | Your AI answer may be seen before your listing. | Whitespark, 2026 |
| Recent reviews carry more weight. | 74% of consumers specifically look for reviews written in the last three months. | A steady stream of new reviews feeds both customers and AI. | BrightLocal, 2026 |
| Technical shortcuts keep disappointing. | Adding JSON-LD schema produced no measurable change in AI citations across 1,885 tracked pages, and 97% of published llms.txt files got zero requests in May 2026. | Clear, current content beats hidden code. | Ahrefs, 2026 |
| Better content still moves the needle. | Researchers from Princeton and Georgia Tech found optimizing content boosted visibility in generative engine responses by up to 40%. | Writing that answers real questions is measurable, not guesswork. | Aggarwal et al. |
What most people miss: the businesses AI recommends aren’t always the biggest. They’re the ones whose information is clear, consistent and recent across the web.
What should you fix first?
Fix the facts customers act on (hours, services, locations and contact details) everywhere they appear, then make your website answer the questions people actually ask.
Start with the listings. Claim or update your Google Business Profile and Bing Places, then correct the three or four directories that rank for your business name. Next, rewrite vague website copy into plain answers: which towns you serve, what a typical job costs, how quickly you respond. Finally, ask recent customers for reviews, because fresh ones count.
If you want a baseline before you start, a free AI visibility audit shows how your site holds up against the criteria AI engines use to pick sources. In my view, the order matters more than the tool: facts first, answers second, reviews third.
Start today: open ChatGPT, type “What does [your business] in [your city] do?” and read the answer as if you were a customer.
Written by
Alex Shortov
CTO, AEO Content
Full-stack engineer and content infrastructure architect with 20 years of building enterprise systems.
Connect on LinkedIn
Frequently Asked Questions: What else do business owners ask about AI answers?
The questions below cover Google reviews, schema markup, llms.txt files, customer trust and how often to check.
Does ChatGPT use my Google reviews?
Not directly. Search Engine Land’s local testing found ChatGPT relies on Bing-indexed web results and its own selection logic. It may reference Google ratings when another crawled page republishes them, so don’t treat Google reviews as your whole strategy.
Will adding schema markup fix wrong AI answers?
Probably not on its own. An Ahrefs study of 1,885 tracked pages found adding JSON-LD schema made no measurable difference in AI citations. Schema still helps search engines, but clear, accurate text does more for AI answers.
Do I need an llms.txt file?
It’s low priority for most small businesses. Ahrefs found 97% of published llms.txt files received zero requests in May 2026 across more than 137,000 domains. Spend that time updating your listings instead.
Do customers trust what AI tells them about businesses?
Many do, but most double-check. BrightLocal’s data shows 88% of consumers verify AI recommendations, so your listings, reviews and website need to agree with each other.
How often should I check what AI says about my business?
I’d run the three-prompt test once a month, and again after any change to your hours, services or location. AI answers shift as new information appears, and ChatGPT in particular favors recent content.
