5 AI Models, 4 Signals: What 120K Mentions Reveal About Multi-Location SEO

This post was sponsored by Uberall. The opinions expressed in this article are the sponsor’s own.

How do I get my locations recommended by ChatGPT, Gemini, or Perplexity?

Do AI models pick local businesses the same way Google does?

Does review volume or star rating matter more for AI visibility?

We analyzed 120,000+ AI mentions across 3,793 locations and five models, from ChatGPT to Perplexity. Each model has its own personality, but they all reward the same signals, and market share isn’t one of them. Here’s what to focus on this month.

In This Report

1. The Differences Between Each AI Model2. The 4 AI Mention Factors3. 3 Things to Do This Month4. Be a Local Role Model for All AI Models

The Differences Between Each AI Model

A mere year ago, the verdict on AI and local search was that it was pretty beta-esque: Local results were still vague, missing images, and recommended places that weren’t even nearby.

Today, AI tools deliver more useful local results, more consumers use them for local recommendations, and we know more about where they pull business information from.

Beyond this, our recent in-house experiments at Uberall tell us each model has its own recommendation “personality.”

Claude is the most conservative. It favors local and community businesses but largely avoids naming specific healthcare providers, likely reflecting caution around anything resembling medical guidance. Gemini is the most diverse, likely because it cross-references live Google Maps data. It surfaces 8× more unique restaurants than ChatGPT in the same study. ChatGPT is consensus-driven and concentrated. It generates a short, sticky shortlist, but with the highest hallucination rates across verticals. Grok it references chef qualifications and Instagram content more than any other model, and its responses read like long-form food journalism: “must-try dishes,” chef names, cooking methods, and historical context. Perplexity searches live, cites its sources, and generates the most mentions per run. For businesses with strong current web presence, it’s the “easiest” model to win on.

Multi-location brands have spent years learning Google’s “personality.” Now there are at least five more to figure out. Just like marketers need to understand their customers, they need to understand what matters to each AI model.

Our Uberall data comes from 120,000+ AI mentions across 3,793 locations in US cities from Chicago to New York, covering dentists, restaurants, grocery stores, hotels, and banks. Our internal GEO analyst Katya Shishchenko looked for patterns consistent across industries and models; you’ve heard of ranking signals and factors, so let’s call these “AI mention factors.”

The 4 AI Mention Factors

I didn’t want to make an acronym, but since we’re personifying AI systems anyway, we might as well just say the models like BARS:

Business data. Authority. Review. Social. Image created by Uberall, September 2026

This article lays out what drives AI visibility across those four categories, before giving you the three things to focus on this month. None of it is exotic. It just needs to be done well, intentionally, and at scale.

Factor 1: Business Data

GBP completeness and richness, along with business age, determine whether a local brand shows up in LLM responses at all, not how often it appears once it’s mentionable.

Business age is obviously not influence-able, but the following are, each with vertical-specific statistics from our study.

A filled-out GBP description: This optimization can mean a three-fold increase in mention rates for grocery stores. Relevant attributes: Going from 6–10 attributes to 31–50 takes mention probability from 22% to 94% in the hotel category. Photo count: It’s the single strongest predictor of restaurant mention frequency. Top-mentioned restaurants average three times more photos. Photo count is also the only signal in dental that predicts both whether a practice gets mentioned and how often. And in banking, it ranks among the three strongest predictors of AI mentions overall. Location count: More locations means higher mention rates, reaching 100% at scale in both grocery and dental categories.

Most of these factors get your brand’s foot in the figurative door rather than determining how often you come up in AI conversations, but that makes them the right place to start.

Factor 2: Authority Signals

When we think about brand authority in multi-location marketing, we typically think of scale: enterprise brands, big footprints, big market share. But across all five verticals, brand size (store count, practice count, deposit share, room supply) was a poor predictor of how often AI models recommend a brand.

Size did influence whether a brand got mentioned at all in grocery, hotels, and banking, where chains benefit from signals LLMs picked up during training. For restaurants and dentists, independent brands can and do outperform chains on mention rates.

Aside from brand size, other authority signals determine how prominently and consistently multi-location brands show up:

Media mention frequency: Brands with 30+ news mentions saw a 15-fold frequency increase in banking and a 100% mention rate in grocery. Editorial list presence: Bankrate, Forbes Travel Guide, and Michelin mentions function almost like direct “ingestion” into LLM training data. Banks on three or more editorial platforms see a 13-fold mention increase; Michelin recognition shows up in 94.5% of Perplexity’s restaurant responses. Wikipedia presence: This is a reliable positive mention factor for hotels, grocery stores, and banks, though not for dentists.

Delivery platform presence, on Instacart, DoorDash, or Mercato, for example, does not influence grocery stores’ AI visibility.

Earned media and editorial authority are still the long game. Our study simply reemphasized this, and the continued importance of E-E-A-T. There’s no faking or shortcutting your way onto a Michelin list, and that’s exactly what makes these signals so powerful once you’ve earned them.

Factor 3: Review Signals

Here’s the finding we think is the big beat drop of this article: Across all five verticals, review volume predicts AI mentions. Star ratings are a weaker AI mention factor in four of five business categories; where they do show up as a signal, they’re never the thing driving visibility.

This is what our study found for three of the business types we analyzed:

Dentists: No star rating on any platform reaches statistical significance. Grocery stores: Brands with higher review volume but lower ratings get mentioned 94.3% of the time, versus 60.6% for high-rated but low-volume brands. Banks: Higher aggregate ratings on Yelp and TrustPilot actually correlate negatively with mention frequency. That’s not because bad reviews help; it’s because the most-mentioned banks are the national giants, and serving millions of customers means accumulating complaints at scale. Hotels: The exception to the pattern. GBP star ratings correlated more strongly with AI mention probability and frequency than review count.

When teams optimize their brand’s online visibility for review volume (while, of course, not dropping their service standards), it matters where they collect and manage their reviews. Which platform matters most depends on the vertical.

While star ratings aren’t always a reliable AI mention factor, they still influence the humans deciding which business to visit. My advice is to optimize review volume for the models, and keep the rating for the people.

Factor 4: Social Signals

Instagram and Facebook are each mention factors, but they play different roles in local AI visibility. Facebook gets brands into the generated response, Instagram gets them talked about once they’re there.

Facebook followers enhance AI mention probability, meaning that business wins the mention in the first place. For banks, follower count is the strongest social AI mention factor overall, and dentist practices that get mentioned have almost five times the number of followers as those that don’t.

Instagram presence amplifies AI mention frequency. Restaurants with a strong Instagram and Yelp presence are mentioned almost seven times more than those that don’t. For boutique hotels specifically, Instagram is the single strongest predictor of AI mentions, outperforming even GBP and editorial signals. And remember Grok’s persona from earlier; it references Instagram content more than any other model.

The takeaway is that regular activity and growing your audience on your social profiles, across locations, is worth the work.

3 Things to Do This Month

Based on what we found, AI models consistently like BARS, even if each model “persona” ranks them slightly differently by business type.

The brands showing up more consistently across AI models are regularly optimizing for all of these mention factors at scale, across every location, every profile, every review platform. It’s what we call Location Performance Optimization (I’ve spoken about this at brightonSEO), and it’s how multi-location brands stop relying on market share and authority alone to show up in local AI search results.

Our advice fits on a bar napkin.

1. Complete Your GBP Profiles

Completing your GBP descriptions, attributes and categories at scale can increase AI mention rates by 40 to 60 percentage points in some verticals. And if your location profiles aren’t complete enough to get mentioned in the first place, it doesn’t matter how many reviews you’ve collected; there’s no local business to amplify.

Our in-platform agentic AI UB-I keeps every location profile complete, fixes outdated ones after human review, and lists profile errors in order of search impact.

Screenshot of UB-I Agent Control Center Dashboard by Uberall, September 2026

2. Prioritize Review Volume Over Star Ratings

Across all five verticals and models, review volume predicts AI mentions better than ratings do. If your team is spending most of its energy fixating on a 4.3 rating, some of that effort is better spent gathering fresh reviews across relevant platforms instead.

3. Increase Your GBP Photo Volume

Photo count is the number-one predictor of restaurant mention frequency, the strongest signal in dental, and the third most important AI mention factor for banks.

Aim to upload 100+ quality (and, ideally, location-specific) photos across your locations. For restaurants and hotels targeting top-tier AI positioning, aim for 2,000+ photos.

A tip: Upload in steady increments rather than all at once. This signals an active, living business.

Be a Local Role Model for All AI Models

Local AI search isn’t in its beta era anymore, and the optimizations teams should prioritize have become figure-out-able. Acronym-able, even.

The way multi-location brands grow their AI visibility is by meeting the models at the BARS: prioritizing long-term data correctness and coherence at scale, building authority that earns media and editorial mentions, consolidating social proof through reviews, and showcasing the experience of visiting them through photos.

Doing all of that across hundreds of locations is a lot of work, which is exactly why we built UB-I.

---

Image Credits

Featured Image: Image by Uberall. Used with permission.

In-Post Images: Images by Uberall. Used with permission.