Brands with hundreds of locations show up in Google's local top three about 36% of the time. ChatGPT recommends those same locations 1.2% of the time.
Those numbers come from SOCi's 2026 Local Visibility Index, which looked at more than 350,000 locations across 2,751 multi-location brands. Neil Patel built a whole video around them, and he ran a live test: he asked ChatGPT for an auto repair shop in Beverly Hills. None of the big national chains came up.
I watched it because the same logic applies to any local business, with one location or four hundred. The gap is where the independent shop wins, and it is also where a chain location can win if someone does the work. Here is what Neil found and what I would do with it.
The gap: 36% on Google, 1.2% on ChatGPT
Ask ChatGPT for a gym near you, a burger place or an auto repair shop. These categories are full of national brands with thousands of locations and decades of advertising. ChatGPT often skips them and recommends the independent shop down the street.
Neil's point is that knowing a brand and recommending one of its locations are two different things. ChatGPT knows Jiffy Lube. That does not mean it will send you to the Jiffy Lube two blocks away.
The SOCi numbers show how different the two channels are. On Google, a strong multi-location brand is in the local top three about a third of the time. On ChatGPT, almost never. Gemini did better in the same study at 11%, still far below Google.
→ Ranking in Google Maps does not carry over to ChatGPT by itself.
Every franchise is two companies
To explain why the big names lose, Neil starts with how franchises are built. From the outside it is one brand: one logo, one menu, one jingle. From the inside it is two businesses with different goals.
Corporate owns the brand and collects a royalty on gross sales. It gets paid on every dollar that goes through the register, whether or not the store made money that month. The local operator owns the store and lives on net profit, what is left after food, labor, rent and the marketing they are required to spend.
So corporate wants volume and the operator wants margin. His example is McDonald's $5 meal deal. The whole country saw it. An equity research firm surveyed franchisees and found sales had moved 1.3%. Corporate got the headlines. The operators paid for the discount out of their own margins.
There is also a time difference. Corporate builds a brand over ten years. The operator has to make payroll on Friday. Neil is clear that nobody is the villain here. Both sides are acting rationally for the model they are paid on.
→ The people who control a chain's marketing are not the people who feel a slow week at one location.
Nobody can see the marketing money
In most franchise systems, the operator pays 1 to 3% of every sale into a national marketing fund that corporate controls. On top of that, 85% of franchise brands require or recommend that operators also spend their own money on local marketing.
Neither side can see much. Almost half of those brands run the local spend requirement on the honor system, with no reporting. When operators ask what happened to their contribution to the national fund, they usually get impressions and reach.
That is now in court. Neil mentions franchise owners at a pizza chain, a burger chain and an auto body franchise suing their own brands over marketing funds. They are not asking for money. They are asking for audit rights and transparent accounting.
→ While both sides argued about the budget, customers changed how they choose a local business.
How ChatGPT picks a local business
Neil tells a story about his mom, who ran a daycare. She got families through conversations at pickup and by knowing every teacher at the local elementary school. Today a parent who hears about a daycare from a friend searches it first, looks at the photos, reads the reviews, checks how the owner answered the angry one, and only then calls.
AI took that habit and scaled it. According to the numbers Neil shares, AI use for local search went from about 6% to about 45% in a year.
The important detail is how ChatGPT searches. Researchers who studied its search behavior found it often writes its own query before fetching anything, and that query can already contain brand names the user never typed. The short list comes from what has been written and said about businesses over the years. Then the search checks those names and can surface others.
Google explicitly uses prominence, meaning how well known a business is, in local rankings. Brand recognition can also influence which names ChatGPT considers. But fame gets a brand into the conversation. It does not guarantee a specific location gets into the answer.
We covered the wider shift in our piece on what AI actually changed in local SEO. Neil's video adds the location-level view.
→ The AI decides at the level of the address, not the logo.
The Beverly Hills test
Neil asked ChatGPT for an auto repair shop near him in Beverly Hills. Auto repair is a category packed with national chains: Firestone, Pep Boys, Jiffy Lube.
ChatGPT's first choice was Beverly Hills Autoworks. The other options were Royal Motors Beverly Hills and AGT Automotive. None of the national brands appeared.
He suggests you run the same test for your own category and city, and try a few variations, since the answers change. The useful question is whether your location shows up when someone nearby asks for what you sell.
→ Run this test for your own business today. It takes two minutes and it is the most honest report you will get.
What each location needs
These are the signals Neil says each location has to get right, one location at a time:
- Star rating. There is a floor at roughly 4.0 stars. Below it, most AI systems filter the location out entirely.
- Review recency. A third of consumers look for reviews written in the past two weeks, and 44% say a review from the past month is one of the things they trust most.
- Response rate. The average business responds to less than half of its Google reviews. 80% of consumers say they are more likely to use a business that responds to every review, and one in five expects a response the same day.
- Photos and posts on that location's profile.
His team also surveyed 500 marketers and business owners on what moves visibility most. Reviews and sentiment came in second, at 91%.
Neil's summary of the split: corporate builds the brand recognition, but accurate hours, services at that address and what customers say in the reviews have to be managed location by location.
→ Every item on this list belongs to the location, not to the brand.
Why letting locations go rogue backfires
The obvious fix would be to let each location run its own marketing. Neil explains why corporate is afraid of that.
First, legal risk. Under federal law, one unsubstantiated claim from one location can trigger action against the whole system, and the FTC can fine violations of the Franchise Rule at over $50,000 per violation. One store posts it, every store carries the risk.
Second, ownership. Most franchise agreements never said who owns the Google Business Profile or the reviews in it. Neil tells the story of a top franchisee who left her system, kept the profile with 100 five-star reviews and pointed it at her new business. The franchisor ended up buying its own Google profile back.
Third, rogue marketing hurts the operator who does it. An off-brand micro site splits authority away from the location's real page, so both rank worse. Inconsistent name, address and phone number across listings make the location look ambiguous to Google, which ranks it lower. A duplicate map listing risks a profile suspension. If you run a single location, this part applies to you too, and our guide to citation building covers how to keep your details consistent.
The brands that get this right change corporate's job from brand police to supply chain. Neil's example is a quick service chain that runs 42 local TikTok accounts instead of one national one. Corporate gives the stores voice guidelines, pre-cleared sounds, caption templates and B-roll. In the pilot stores, the slowest afternoon window saw a 19% lift in transactions with no national ad spend.
→ Make the compliant option the easiest option for every location.
It is a systems problem
Neil lists what winning the AI layer actually requires: every review answered, at every location, forever. A rating above 4.0 at every location. Fresh reviews all the time. Attribution by territory. Name, address and phone number matching across every directory.
His point is that none of these is a creative problem. All of them are systems problems. At four locations you can get it done with goodwill and a shared drive. At 400 locations, either a system does it or it does not happen. One of the brands his agency runs has 400 locations, which means 400 profiles and 400 review streams.
This is also true at one location. The owner of a single dental practice or plumbing company does not skip review replies because they don't care. They skip them because it is Thursday night and they have been working since 7am.
That is what we built Distribb for local businesses to handle. You connect your Google Business Profile once, set the area you serve, and then it runs:
- A reply drafted for every review, written to match the rating and what the customer actually said, so nothing sits unanswered for weeks.
- Google Business Profile posts on a schedule you set once, with a link and an image where they help.
- Location-aware articles: keyword research targeted to your region, and content that names the places you serve and the work you do there.
- One project per branch, each with its own keywords, content and profile connection, so locations never overwrite each other.
- Profile performance in one place: views, review counts and rating movement.
- Backlinks from real businesses through the exchange, running alongside the profile work.
Pro is $97 a month per project with 30 articles. Accelerator is $495 a month with 90 articles, done-for-you videos and a human on our team reviewing every piece. Both come with a 3-day free trial.
The three things to put in writing
Technology is only part of Neil's answer. For chains, franchised or company-owned, he says corporate and local teams do not need to agree on a strategy. They need to agree on three things, in writing:
- What each side is accountable for.
- What each side is allowed to see.
- The one number both are judged on. His choice: how often AI names this location when someone nearby asks. Both sides can see it and neither side can fake it.
With that in place, the operator gets permission and proof, and the brand shows up locally without losing control. I would add a fourth line to any franchise agreement after hearing the story about the franchisee who walked away with the profile: who owns the Google Business Profile and its reviews.
→ Agree on the number first. The arguments about budget get shorter after that.
What I'm taking from this
The independent shop has a real opening. A 1.2% recommendation rate for chain locations means ChatGPT's answers are mostly going to someone else. If you run a local business, that someone can be you.
Reviews are the work. Rating above 4.0, recent reviews, and a reply to every single one. Nothing else on the list moves as much.
Consistency beats cleverness. The same name, address and phone number everywhere, one profile per location, no side micro sites.
Test it yourself every month. Ask ChatGPT for your category in your city, a few different ways, and write down who it names.
Put a system on it. Whether you have one location or 400, the tasks repeat forever. Target your content with keyword research for local SEO, then let software handle the recurring parts.
Watch the full video
Neil's video is 16 minutes. The live ChatGPT test starts a little after the middle.
FAQ
Why doesn't ChatGPT recommend big chains?
Because it recommends locations, not brands. SOCi's 2026 index found chain locations in Google's local top three about 36% of the time but recommended by ChatGPT only 1.2% of the time. ChatGPT weighs each location's rating, recent reviews, review replies and consistent details, and chain locations are often weak on those.
What star rating do you need to be recommended by AI?
Roughly 4.0 stars or higher, according to Neil Patel. Below that floor, most AI systems filter a location out instead of ranking it lower. SOCi's data also shows locations recommended by ChatGPT average about 4.3 stars, so a rating well above 4.0 gives you more room.
Do review replies help with ChatGPT visibility?
They help with the customers who read the reviews first, and responsiveness is one of the location signals Neil lists. The average business answers less than half of its Google reviews, while 80% of consumers say they prefer a business that responds to every one, and one in five expects a reply the same day.
Who owns a franchise's Google Business Profile?
Often the agreement does not say, which is the problem. Neil describes a franchisee who left, kept a profile with 100 five-star reviews and pointed it at her new business, so the franchisor had to buy it back. Put profile and review ownership in the franchise agreement explicitly.
How do I check if ChatGPT recommends my business?
Ask it directly. Type your category and city the way a customer would, for example "auto repair shop near me in Beverly Hills", try a few variations, and note which businesses it names. Repeat monthly for each location, since answers change from one run to the next.
Get each location found
The businesses ChatGPT recommends are the ones with fresh reviews, replies to all of them, and pages that talk about the places they serve. If you want that to keep happening without it landing on your evenings, see how Distribb runs local SEO for each location.