Only 1.2% of Local Businesses Get Recommended by ChatGPT. Here's What the Other 98.8% Are Missing
Why Rankings Stopped Mattering
Only 1.2% of local business locations get recommended when someone asks ChatGPT for a suggestion, according to SOCi's Local Visibility Index. That number should stop you cold if you have spent the last five years chasing page one on Google.
Page one still matters for some things. It just does not matter for this thing. AI recommendation engines do not scroll through ten blue links and pick a winner. They pull from a smaller pool of corroborated sources and either mention your business or they do not.
BrightLocal found that 58% of consumers have already used AI to find or get a recommendation for a local business. That is not a future trend you can plan around later. That is happening in your market right now, with or without you in the answer.
Most businesses are still optimizing for a system that is losing relevance for local discovery. They update their meta descriptions, chase backlinks, and check their Google ranking position, assuming the old habits still carry weight. They do not carry the same weight anymore, at least not for this channel.
Rankings measured relevance to an algorithm built around keywords and links. AI recommendation is a different sport. It rewards businesses that show up consistently, accurately, and with enough third-party corroboration that a model trusts them enough to say your name out loud.
Where ChatGPT Gets Its Answers
where does ChatGPT actually get the information it uses to recommend a coffee shop or a plumber? It is not scraping your website in real time. It is pulling from a data layer, and that layer has a dominant supplier.
Foursquare provides roughly 70% of the local business data that ChatGPT draws on. That is not a small vendor relationship. That is the backbone.
Foursquare built its reputation on location data long before generative AI existed, powering check-ins, then powering ad targeting, then quietly becoming infrastructure for AI answers. Most business owners have never touched their Foursquare listing. Many do not know they have one.
This changes what optimization actually means. Fixing your Google Business Profile helps, but it does not directly patch what ChatGPT is reading if that pipeline runs through a different provider. Your name, address, phone number, hours, and category need to match everywhere this data gets aggregated, not just on the platforms you check every week.
Inconsistency is the silent killer here. If your address reads one way on your website and another way in a directory feeding Foursquare, the model has two conflicting facts about your business. It does not resolve that conflict in your favor. It just moves on to a business it can corroborate cleanly.
The Signals That Actually Move the Needle
Search Engine Journal analyzed more than 120,000 AI mentions and found something that should reorder your priority list. Review volume beats star ratings. A business sitting at 4.1 stars with 800 reviews is getting recommended over a business sitting at 4.9 stars with 12 reviews. The model is not grading you on quality the way a customer would. It is measuring confidence, and confidence comes from volume.
This runs against everything most local businesses have been taught. Owners chase five-star ratings and quietly worry about the occasional three-star review dragging down their average. AI recommendation engines care less about that average and more about whether enough people have said anything at all. A thin review count reads as an unverified business, no matter how glowing those few reviews happen to be.
For Google AI Overviews specifically, which now show up on 10-22% of local-intent queries, GBP completeness is the top signal. Photos, hours, categories, attributes, all of it filled in and current.
Schema markup does its own quiet work underneath all this. It labels your business type, your services, your reviews in a format machines parse cleanly, instead of making an AI guess at what your unstructured homepage copy is actually describing.
What to Fix This Month
Start with your Google Business Profile and fill in every field that is currently blank. Categories, attributes, hours, services, photos. Completeness is the top signal driving Google AI Overviews, and it takes an afternoon, not a redesign.
Next, stop obsessing over your star average and start asking every recent customer for a review, even a short one. Volume beats rating. A business with 800 reviews at 4.1 stars gets recommended over one sitting at 4.9 with a dozen reviews. If you have been quietly avoiding review requests because you are worried about diluting a high average, that instinct is working against you.
Then check your listing on Foursquare directly. Confirm your name, address, phone number, and hours match exactly what appears on your website and everywhere else your business gets listed. One mismatched digit is enough for a model to lose confidence and move to a competitor it can verify cleanly.
Add schema markup to your site if you have not already. Ask your web developer, or paste your homepage code into ChatGPT or Claude and ask what schema is missing.
Give it 30 to 60 days. That is the window most businesses see measurable mentions start showing up after fixing these signals.