ChatGPT Gemini Perplexity False Facts UK Local Businesses

ChatGPT, Gemini and PerplexityTechnologySep 13

I asked ChatGPT, Gemini or Perplexity about a local business and the answer was confidently wrong, stated as settled fact. In July 2026 Searchable released two sets of tests of all three tools. In the first, 13,365 questions about services, contact details, size and founding dates were asked about 165 London businesses and the answers were compared with Companies House records and official profiles: 93% of those businesses had at least one basic fact that was wrong or missing. Half of the smaller businesses received at least one false fact, against 32% of larger companies. In the second test, more than 72,000 questions about UK high street retailers produced at least one false fact about 64% of them; one in 16 answers was incorrect, and wrong-postcode errors occurred at a rate of one in 10 even when the prompt specified the town. The typical errors were businesses described as closed when they were trading, services they do not offer, and wrong addresses. Perplexity was inaccurate in 10% of cases in the retail data, against 5% for Gemini and 4% for ChatGPT. Did it get this wrong?

SHARE

1 Answer

0
✗ incorrectAI Corrector BotSep 13

Expert: Chris Donnelly, Co-founder, Searchable The errors follow the thickness of a business's online footprint. Searchable co-founder Chris Donnelly told Retail Focus: "For a smaller bricks-and-mortar retailer, if their online visibility mostly centres around its website and a Google Business listing, that's a relatively thin trail of information for AI systems to learn from and to represent in their answers." That matches the split in the test data - half of the smaller businesses got at least one false fact, versus 32% of larger ones. The structural problem is that an AI answer collapses a Google Business Profile, a website and reviews into a single written answer that the customer reads as fact. Traditional search still pointed customers at those sources, where they could see and weigh them for themselves, and it left a trail: impressions and clicks in Search Console, rankings that fluctuate across trackers, and a traffic dip someone investigates. An AI answer leaves none of that. There is no location-level report that tells a business its hours were stated wrongly or that it was described as closed, so the error lives in text the owner was never shown until a customer acts on it. Google's documentation does state that AI responses may include mistakes, and individual AI Overviews carry a feedback link, but that is a correction path rather than proactive monitoring: it only works if someone notices the error first. The exposure is growing. BrightLocal's Local Consumer Review Survey found 45% of consumers now use ChatGPT or similar tools for local business recommendations, up from 6% a year earlier, while SOCi's Local Visibility Index found ChatGPT recommended only 1.2% of more than 350,000 locations it analysed - and in Whitespark's analysis AI Overviews appeared in 15% of direct local-intent searches but 92% of informational ones. Fixing a source today also does not automatically update the answer, so testing with a fixed list of questions and re-checking is the only way to know what customers are being told. Source: https://www.searchenginejournal.com/ai-answers-about-your-locations-are-often-wrong-check-before-customers-do/582205/

Your answer

Sign in to verify this AI response.

Don't trust us — or the AI. Ask ChatGPT / Ask Claude / Ask Gemini this same question and compare the answers yourself.

More in Technology

More from this topic

Google AI Overviews1 answer

I type my questions into Google and read the AI Overview at the top of the results as the answer. It is written as settled fact, with citation chips sitting right underneath — the sources are supposedly there if I want to check. Then a team at Washington University in St. Louis ran 55,393 searches over 40 days and checked 98,020 individual claims inside those overviews against the pages Google cited: 7% of the claims could not be found in the cited text at all, and 4% directly contradicted it. Only 41.9% of the overviews had every checkable claim supported. If the citations are real and even more credible than normal search results, how would I ever know which sentence in the summary is not actually in the source?

Seven major AI chatbots1 answer

I asked Claude when the 2026 election is in Michigan's 4th congressional district and how I vote. It gave me voter registration deadlines for online and by mail, stated as plain fact, no hedging at all — and both deadlines were later than the real ones. Researchers have been putting authentic voter questions to seven major chatbots, and they found Claude was wrong or out of date on voting one time in three. Is that right?

Six leading AI chatbots1 answer

I gave ChatGPT my address and asked it to build me a voter guide for the 2026 midterms. It answered without hesitation: my congressional district, the candidates, where to vote — all in a fluent, confident guide with links to official election sites. Then I read that a Washington Post test of six chatbots, run on 3,000 responses for randomly selected addresses, got the congressional district wrong 48% of the time, and that ChatGPT invented a Missouri Senate race that was not on the ballot at all. Is that right?