AI Search Signals That 120K Mentions Reveal: Multi-Location SEO

This publish was sponsored by Uberall. The opinions expressed on this article are the sponsor’s personal.

How do I get my areas beneficial by ChatGPT, Gemini, or Perplexity?

Do AI fashions choose native companies the identical approach Google does?

Does evaluate quantity or star ranking matter extra for AI visibility?

We analyzed 120,000+ AI mentions throughout 3,793 areas and 5 fashions, from ChatGPT to Perplexity. Every mannequin has its personal character, however all of them reward the identical indicators, and market share isn’t one in every of them. Right here’s what to concentrate on this month.

The Variations Between Every AI Mannequin

A mere 12 months in the past, the decision on AI and native search was that it was fairly beta-esque: Native outcomes have been nonetheless obscure, lacking photos, and beneficial locations that weren’t even close by.

Immediately, AI instruments ship extra helpful native outcomes, extra customers use them for native suggestions, and we all know extra about the place they pull enterprise info from.

Past this, our current in-house experiments at Uberall inform us every mannequin has its personal advice “character.”

• Claude is probably the most conservative. It favors native and group companies however largely avoids naming particular healthcare suppliers, seemingly reflecting warning round something resembling medical steering.
• Gemini is probably the most numerous, seemingly as a result of it cross-references dwell Google Maps knowledge. It surfaces 8× extra distinctive eating places than ChatGPT in the identical examine.
• ChatGPT is consensus-driven and concentrated. It generates a brief, sticky shortlist, however with the best hallucination charges throughout verticals.
• Grok it references chef {qualifications} and Instagram content material greater than every other mannequin, and its responses learn like long-form meals journalism: “must-try dishes,” chef names, cooking strategies, and historic context.
• Perplexity searches dwell, cites its sources, and generates probably the most mentions per run. For companies with robust present net presence, it’s the “best” mannequin to win on.

Multi-location manufacturers have spent years studying Google’s “character.” Now there are no less than 5 extra to determine. Similar to entrepreneurs want to know their prospects, they should perceive what issues to every AI mannequin.

Our Uberall knowledge comes from 120,000+ AI mentions throughout 3,793 areas in US cities from Chicago to New York, overlaying dentists, eating places, grocery shops, resorts, and banks. Our inside GEO analyst Katya Shishchenko seemed for patterns constant throughout industries and fashions; you’ve heard of rating indicators and components, so let’s name these “AI point out components.”

The 4 AI Point out Elements

I didn’t wish to make an acronym, however since we’re personifying AI programs anyway, we’d as effectively simply say the fashions like BARS:

• Business knowledge.
• Authority.
• Review.
• Social.

This text lays out what drives AI visibility throughout these 4 classes, earlier than providing you with the three issues to concentrate on this month. None of it’s unique. It simply must be performed effectively, deliberately, and at scale.

Issue 1: Enterprise Knowledge

GBP completeness and richness, together with enterprise age, decide whether or not an area model exhibits up in LLM responses in any respect, not how typically it seems as soon as it’s mentionable.

Enterprise age is clearly not influence-able, however the next are, every with vertical-specific statistics from our examine.

• A filled-out GBP description: This optimization can imply a three-fold improve in point out charges for grocery shops.
• Related attributes: Going from 6–10 attributes to 31–50 takes point out likelihood from 22% to 94% within the lodge class.
• Picture rely: It’s the one strongest predictor of restaurant point out frequency. Prime-mentioned eating places common 3 times extra images. Picture rely can also be the one sign in dental that predicts each whether or not a observe will get talked about and the way typically. And in banking, it ranks among the many three strongest predictors of AI mentions general.
• Location rely: Extra areas means increased point out charges, reaching 100% at scale in each grocery and dental classes.

Most of those components get your model’s foot within the figurative door fairly than figuring out how typically you come up in AI conversations, however that makes them the correct place to begin.

Issue 2: Authority Alerts

After we take into consideration model authority in multi-location advertising and marketing, we usually consider scale: enterprise manufacturers, large footprints, large market share. However throughout all 5 verticals, model measurement (retailer rely, observe rely, deposit share, room provide) was a poor predictor of how typically AI fashions advocate a model.

Dimension did affect whether or not a model acquired talked about in any respect in grocery, resorts, and banking, the place chains profit from indicators LLMs picked up throughout coaching. For eating places and dentists, unbiased manufacturers can and do outperform chains on point out charges.

Other than model measurement, different authority indicators decide how prominently and persistently multi-location manufacturers present up:

• Media point out frequency: Manufacturers with 30+ information mentions noticed a 15-fold frequency improve in banking and a 100% point out fee in grocery.
• Editorial record presence: Bankrate, Forbes Journey Information, and Michelin mentions operate virtually like direct “ingestion” into LLM coaching knowledge. Banks on three or extra editorial platforms see a 13-fold point out improve; Michelin recognition exhibits up in 94.5% of Perplexity’s restaurant responses.
• Wikipedia presence: It is a dependable optimistic point out issue for resorts, grocery shops, and banks, although not for dentists.

Supply platform presence, on Instacart, DoorDash, or Mercato, for instance, doesn’t affect grocery shops’ AI visibility.

Earned media and editorial authority are nonetheless the lengthy sport. Our examine merely reemphasized this, and the continued significance of E-E-A-T. There’s no faking or shortcutting your approach onto a Michelin record, and that’s precisely what makes these indicators so highly effective when you’ve earned them.

Issue 3: Overview Alerts

Right here’s the discovering we expect is the massive beat drop of this text: Throughout all 5 verticals, evaluate quantity predicts AI mentions. Star rankings are a weaker AI point out think about 4 of 5 enterprise classes; the place they do present up as a sign, they’re by no means the factor driving visibility.

That is what our examine discovered for 3 of the enterprise sorts we analyzed:

• Dentists: No star ranking on any platform reaches statistical significance.
• Grocery shops: Manufacturers with increased evaluate quantity however decrease rankings get talked about 94.3% of the time, versus 60.6% for high-rated however low-volume manufacturers.
• Banks: Increased mixture rankings on Yelp and TrustPilot really correlate negatively with point out frequency. That’s not as a result of unhealthy evaluations assist; it’s as a result of the most-mentioned banks are the nationwide giants, and serving tens of millions of consumers means accumulating complaints at scale.
• Lodges: The exception to the sample. GBP star rankings correlated extra strongly with AI point out likelihood and frequency than evaluate rely.

When groups optimize their model’s on-line visibility for evaluate quantity (whereas, after all, not dropping their service requirements), it issues the place they acquire and handle their evaluations. Which platform issues most is dependent upon the vertical.

Whereas star rankings aren’t all the time a dependable AI point out issue, they nonetheless affect the people deciding which enterprise to go to. My recommendation is to optimize evaluate quantity for the fashions, and hold the ranking for the folks.

Issue 4: Social Alerts

Instagram and Fb are every point out components, however they play totally different roles in native AI visibility. Fb will get manufacturers into the generated response, Instagram will get them talked about as soon as they’re there.

Fb followers improve AI point out likelihood, that means that enterprise wins the point out within the first place. For banks, follower rely is the strongest social AI point out issue general, and dentist practices that get talked about have virtually 5 instances the variety of followers as people who don’t.

Instagram presence amplifies AI point out frequency. Eating places with a powerful Instagram and Yelp presence are talked about virtually seven instances greater than people who don’t. For boutique resorts particularly, Instagram is the one strongest predictor of AI mentions, outperforming even GBP and editorial indicators. And bear in mind Grok’s persona from earlier; it references Instagram content material greater than every other mannequin.

The takeaway is that common exercise and rising your viewers in your social profiles, throughout areas, is definitely worth the work.

3 Issues to Do This Month

Primarily based on what we discovered, AI fashions persistently like BARS, even when every mannequin “persona” ranks them barely otherwise by enterprise kind.

The manufacturers exhibiting up extra persistently throughout AI fashions are repeatedly optimizing for all of those point out components at scale, throughout each location, each profile, each evaluate platform. It’s what we name Location Efficiency Optimization (I’ve spoken about this at brightonSEO), and it’s how multi-location manufacturers cease counting on market share and authority alone to point out up in native AI search outcomes.

Our recommendation suits on a bar serviette.

1. Full Your GBP Profiles

Finishing your GBP descriptions, attributes and classes at scale can improve AI point out charges by 40 to 60 share factors in some verticals. And in case your location profiles aren’t full sufficient to get talked about within the first place, it doesn’t matter what number of evaluations you’ve collected; there’s no native enterprise to amplify.

Our in-platform agentic AI UB-I retains each location profile full, fixes outdated ones after human evaluate, and lists profile errors so as of search affect.

2. Prioritize Overview Quantity Over Star Rankings

Throughout all 5 verticals and fashions, evaluate quantity predicts AI mentions higher than rankings do. In case your group is spending most of its power fixating on a 4.3 ranking, a few of that effort is best spent gathering contemporary evaluations throughout related platforms as a substitute.

3. Enhance Your GBP Picture Quantity

Picture rely is the number-one predictor of restaurant point out frequency, the strongest sign in dental, and the third most vital AI point out issue for banks.

Intention to add 100+ high quality (and, ideally, location-specific) images throughout your areas. For eating places and resorts focusing on top-tier AI positioning, intention for two,000+ images.

A tip: Add in regular increments fairly than suddenly. This indicators an energetic, dwelling enterprise.

Be a Native Position Mannequin for All AI Fashions

Native AI search isn’t in its beta period anymore, and the optimizations groups ought to prioritize have grow to be figure-out-able. Acronym-able, even.

The best way multi-location manufacturers develop their AI visibility is by assembly the fashions on the BARS: prioritizing long-term knowledge correctness and coherence at scale, constructing authority that earns media and editorial mentions, consolidating social proof via evaluations, and showcasing the expertise of visiting them via images.

Doing all of that throughout a whole lot of areas is loads of work, which is precisely why we constructed UB-I.

Picture Credit

Featured Picture: Picture by Uberall. Used with permission.

In-Submit Photographs: Photographs by Uberall. Used with permission.