Only 1.2% of Local Businesses Get Recommended by ChatGPT. Here's Why the Other 98.8% Are Invisible.

A landmark study of 350,000 business locations reveals that AI-powered discovery is 30 times more selective than Google. If you're not in that 1.2%, your customers are being sent to someone else.

There's a number circulating right now that should fundamentally change how every restaurant, hotel, retailer, and service business thinks about their online presence. SOCi's 2026 Local Visibility Index, the largest study of its kind, analyzed over 350,000 business locations across 2,700 enterprise brands. The finding that matters most: only 1.2% of those locations were recommended by ChatGPT.

Not 12%. Not even 5%. Just 1.2%.

For context, the same businesses appeared in Google's local 3-pack 35.9% of the time. That means getting recommended by ChatGPT is roughly 30 times harder than ranking in traditional local search. And ChatGPT isn't the only AI assistant making these calls. Gemini recommended 11% of locations and Perplexity recommended 7.4%, but the pattern is the same across every platform: AI is radically more selective than Google ever was.

This isn't a minor shift in search dynamics. This is a complete rewrite of how local businesses get found.

AI Doesn't Rank You. It Picks You or Ignores You.

Here's the fundamental difference between Google and AI-powered discovery. Google gives you a page of results. Ten blue links. A local 3-pack. Maybe an ad or two. The user scans, compares, and decides. You might be result number three, but you're still visible.

AI assistants don't work that way. When someone asks ChatGPT "what's the best Italian restaurant near me for a family dinner tonight," it doesn't return a list of ten options. It recommends one, maybe two. There's no second page. There's no "also consider." You're either the answer, or you don't exist in that interaction.

This is why the 1.2% number is so devastating. It means that for every 100 local businesses in a given category, ChatGPT is steering customers toward just one or two of them. The other 98 are invisible, not because they're bad businesses, but because the AI didn't have enough structured, trustworthy data to confidently recommend them.

What Gets You Into the 1.2%

The SOCi study reveals clear patterns in what separates the recommended from the invisible. It's not luck, and it's not just about having a website. Three factors dominate.

Reviews Are Now a Gate, Not a Ranking Signal

In traditional SEO, reviews help your ranking. In AI-powered discovery, reviews determine whether you're eligible to be recommended at all. Locations recommended by ChatGPT averaged 4.3-star ratings. Gemini's recommended locations averaged 3.9 stars. Perplexity's averaged 4.1.

The distinction matters. AI platforms aren't using reviews to rank you against competitors. They're using reviews as a trust filter. If your average rating falls below a certain threshold, you're simply excluded from the recommendation pool entirely. You're not ranked lower. You're not shown at all.

This changes the game for review management. It's no longer about accumulating a high volume of reviews to outrank a competitor. It's about maintaining a consistently high sentiment that keeps you above the AI's trust threshold. A business with 200 reviews at 4.4 stars is more likely to be recommended than one with 2,000 reviews at 3.8 stars.

Data Accuracy Is Embarrassingly Bad, and That's Your Opportunity

Here's a stat that reveals just how early we are in this shift: business profile information was only about 68% accurate on ChatGPT and Perplexity. Gemini scored better at near-100% accuracy, but that's because it pulls directly from Google Maps data.

What does 68% accuracy mean in practice? It means ChatGPT might tell a customer your restaurant is on the wrong street. It might list opening hours from two years ago. It might say you serve cuisine you don't actually offer. Every inaccuracy erodes trust, and trust is the currency of AI recommendations.

But flip that around: if most business data in AI systems is unreliable, then a business that actively ensures its data is correct, consistent, and comprehensive across every platform has a massive advantage. You're not competing against perfection. You're competing against a field where a third of the data is wrong.

Why an Address Isn't Enough: The Hyper-Local Data Gap

There's a deeper layer to the data accuracy problem that most businesses, and most coverage of this report, completely miss. Having your correct address, phone number, and opening hours in structured data is necessary, but it only tells the AI where you are. It doesn't tell the AI what it's like to be where you are.

Think about how real people ask AI assistants for recommendations. They don't say "find me a hotel at this address." They say "find me a beachfront hotel in Split with restaurants within walking distance, near the old town, with good public transport connections." That query requires the AI to understand not just your location, but your surroundings. What's nearby. What's accessible. What the local context actually looks like on the ground.

This is where the vast majority of businesses fall short. Their structured data describes the business in isolation: name, address, category, hours. But AI agents increasingly need contextual location intelligence to make confident recommendations. Which points of interest are nearby? What's the walkability like? Are there verified amenities in the surrounding area? How does the physical environment match what the user is looking for?

Businesses that enrich their structured data with verified, hyper-localized context, the kind of granular surrounding data that lets an AI confidently say "yes, this hotel is 200 meters from the beach, there are six restaurants within a five-minute walk, and the nearest bus stop connects to the airport in 25 minutes," give AI systems exactly the confidence signal they need to make a recommendation. Generic schema tells the AI you exist. Hyper-localized, enriched data tells the AI what it's actually like to choose you.

This is still a blind spot for almost every business. The tools and data sources to generate this kind of surrounding-verified, location-enriched structured data exist, but adoption is barely in its infancy. Which means the window for first movers is wide open, and the competitive advantage for early adopters is enormous. Run a free AEO audit on your business to see exactly which signals you're missing.

Traditional Local Search Dominance Doesn't Transfer

This is perhaps the most surprising finding from the SOCi study. Being visible in Google's local results does not guarantee you'll be visible in AI recommendations. In retail, only 45% of the most visible brands in traditional local search were also frequently recommended by AI platforms.

That means more than half of the businesses dominating Google's local pack are invisible to ChatGPT, Gemini, and Perplexity. The skills, strategies, and investments that made you successful in traditional local SEO are necessary but not sufficient for AI discovery. It's a different game with different rules, and being good at the old game doesn't automatically make you competitive in the new one.

The Industry Impact: Restaurants, Retail, Hotels, and Beyond

The SOCi study covered five key industries: Retail, Food, Financial Services, Local Services, and Property. The 1.2% ChatGPT recommendation rate applies across all of them, which means this isn't an industry-specific quirk. It's a structural feature of how AI-powered discovery works.

For restaurants and food businesses, the implications are immediate. These are high-intent, high-frequency local searches. "Where should we eat tonight" is one of the most common queries to AI assistants. If your restaurant isn't in that 1.2%, you're missing out on a growing stream of customers who are letting the AI decide for them.

For hotels and hospitality, the stakes are even higher because of transaction value. A family choosing a hotel for a week-long holiday based on an AI recommendation represents thousands in revenue. The AI is making that recommendation based on structured data, reviews, and data accuracy, not on how beautiful your website's hero image is.

For retail, the 45% overlap gap between traditional local search visibility and AI visibility means that even established retail brands can't assume their Google presence protects them. AI platforms are evaluating different signals, and many retailers haven't adapted.

For financial services and local services, trust signals carry extra weight. AI platforms are particularly cautious about recommending businesses in categories where a bad recommendation could cause financial or personal harm. The bar for data quality, review sentiment, and consistency is higher, not lower, in these categories.

Why AI Is So Much More Selective Than Google

Understanding why the 1.2% number exists helps explain what you need to do about it. Google can afford to show ten results because the user is expected to evaluate and choose. The cost of showing a mediocre result in position seven is low, since the user will probably skip it.

AI assistants operate under a completely different set of constraints. When ChatGPT recommends a restaurant, it's putting its credibility on the line. A bad recommendation damages user trust in the entire platform. So AI systems apply a much higher confidence threshold before making any recommendation. If the system isn't highly confident that a business will deliver a good experience, it simply doesn't recommend it.

This confidence comes from data signals: review volume and sentiment, data accuracy across platforms, structured business information, recency of updates, and consistency across sources. The AI needs to be sure. And for 98.8% of local businesses, the available data doesn't give it enough confidence to make the call.

The Window Is Open, But It Won't Stay Open

The 1.2% figure represents today's reality, but it also represents today's opportunity. The reason so few businesses are recommended isn't that AI has impossibly high standards. It's that the vast majority of businesses haven't done the work to make their data AI-ready.

Most local businesses still manage their online presence the way they did five years ago: a Google Business Profile that gets updated when someone remembers, a scattering of reviews that nobody responds to systematically, business information that's slightly different on every platform, and a website that talks to humans but is nearly unreadable to machines.

The businesses that fix these fundamentals now, that ensure their data is accurate, their reviews are actively managed, their structured information is comprehensive and current, will move into that 1.2% while their competitors are still wondering why their foot traffic is declining. Our complete AEO guide for local businesses walks through every signal you need to get right.

And here's the compounding dynamic: once an AI learns to trust and recommend your business, every positive customer interaction generates more reviews, which strengthens your position, which leads to more recommendations. It's a flywheel. Getting in early means the flywheel spins longer and faster before your competitors even start pushing theirs.

What This Means for Your Business

The SOCi study makes the situation undeniably clear. AI-powered discovery is not a future concern. It's happening right now, across every industry, in every local market. The platforms are live. The users are there. And the vast majority of businesses are invisible to them.

The 1.2% number isn't meant to discourage you. It's meant to wake you up. Because right now, making it into that group doesn't require breakthrough technology or massive budgets. It requires accurate data, strong reviews, structured information, and the discipline to maintain it all consistently.

That's a solvable problem. But it's only solvable if you recognize it exists, and if you start before your competitors do.

The businesses that move now will define who gets recommended for the next decade. The rest will watch from the sidelines, wondering why the phone stopped ringing.

98.8% of local businesses are invisible to AI assistants. The data to change that already exists inside your business. The question is whether you'll structure it for machines before your competition does.

Related reading:

• What is AEO? Answer Engine Optimization explained
• How AI finds your website in 2026
• NAP consistency: the silent killer of AI visibility
• Check your AI visibility score for free