45% of People Now Use AI to Find Local Businesses. Last Year It Was 6%.
BrightLocal surveyed 1,002 US consumers in February 2026 and found 45% now use ChatGPT and similar tools to find local businesses, up from 6% the year before. That makes AI the third most common way people find a local company, ahead of Yelp and Tripadvisor. Most owners have no idea whether they show up there.
Founder, Simmons Solutions. Three years hands-on with AI.
In plain terms: Almost half of people now ask an AI chatbot for local business recommendations, up from almost nobody a year ago. It has already passed Yelp. When someone asks for a business like yours, the AI names a few companies, and there is no way to buy your way onto that list. This explains how it picks and what actually influences it.
Someone in your area asked ChatGPT for a company like yours last night. It gave them a few names.
That is not a hypothetical anymore. BrightLocal surveyed 1,002 US adults in February and found that 45% now use ChatGPT and similar AI tools to find local businesses, up from 6% a year earlier (survey). It is now the third most common way people find a local company, behind Google at 71% and Facebook at 52%, and ahead of Apple Maps, Yelp at 26%, and Tripadvisor at 24%.
A jump from 6% to 45% in twelve months is not a trend line. It is a channel appearing out of nowhere and passing incumbents that took twenty years to build.
The part that should get your attention
You cannot buy placement in that answer. There is no ad slot, no sponsored tier, no rep to call. When someone asks an assistant for a plumber or a compounding pharmacy or a fence company in your city, it assembles an answer from what it can find written about businesses like yours, and then it names two or three of them.
Everyone else it does not name might as well not exist for that customer. There is no page two.
How it decides, in plain terms
Nobody outside those companies knows the exact recipe, and anyone who tells you they do is guessing. But the inputs are not mysterious, because they are the same signals a careful human would use if you asked them to recommend someone:
Does your information agree with itself? Your name, address, phone number, and hours, matching across your site, your Google profile, and the directories you forgot you were listed in. Contradictions make you a risky recommendation, and a model that is unsure will just name someone else.
Do other people describe what you do? This is the uncomfortable one. An AI answer summarizes what the internet says about you, not what you say about yourself. Reviews, articles, mentions, a supplier page that lists you. A business with no third-party footprint is nearly invisible to a system built on reading what others wrote.
Does your site answer the actual question? People do not ask "best HVAC company." They ask "who can fix a heat pump on a Sunday in Tempe." Pages that answer specific real questions in plain language get pulled into answers. Pages that say "quality service since 1998" do not.
What I would actually do about it
Nothing exotic, and nothing that requires a new budget line.
Go ask. Open ChatGPT and ask for a business like yours in your city, the way a customer would phrase it. Then do it again with two other phrasings. That five-minute exercise tells you more than any report, and most owners have never once done it.
Fix the contradictions first. Same name, same phone, same hours, everywhere. It is boring and it is the cheapest thing on this list.
Get described by someone other than you. Reviews you actually reply to, a mention in a local piece, a partner who lists you. This is the slowest input and the one that compounds hardest.
The honest caveat: nobody can promise you a spot in an AI answer, and I would not trust anyone who does. What you can do is stop being invisible to the thing doing the recommending, which is a different and much more achievable goal.
One more number worth sitting with. Yelp took two decades to become a place people check. This took one year. Whatever you decide to do about it, decide faster than you normally would.
If you want to see what this looks like wired into a business rather than described, the systems page is the plain-English version.
Sources
Keep reading
- Briefing1 in 4 Small Businesses Actually Runs AI. Twice That Many Are Stuck Exploring It.Read
- Briefing42% of Companies Walked Away From Most of Their AI Projects This Year. Last Year It Was 17%.Read
- Briefing95% of Enterprise AI Pilots Show No Measurable Profit Impact. The Tool Was Never the Problem.Read
Get the next one
If this was useful, get one a week. The single AI development that actually matters for operators, in plain English with sources. Free, unsubscribe in one click.