What AI Search Changes for Franchise and Multi-Location Brands, and What It Doesn't

Key takeaways

• AI search rewards your existing search fundamentals rather than a separate discipline you buy: Google says its own AI features are “rooted in our core Search ranking and quality systems,” and ChatGPT, Perplexity, and Gemini each add their own retrieval on top.
• Google says plainly you do not need special AI files, content broken into fragments, or pages rewritten for machines to appear in Google Search, and there is no evidence these tactics earn a universal advantage across AI platforms.
• A network has to keep two things strong at once: one brand entity, and a real, distinct entity for every location, because the AI answers a local query from local signals.
• Google Business Profile, accurate local pages, and genuine reviews feed local prominence, and “there's no way to request or pay for a better local ranking on Google.”
• Fake and AI-generated reviews are banned in the United States under a rule in effect since October 21, 2024, and treated as deceptive marketing in Canada, so the review shortcut is now a liability (U.S. FTC, Rule on Consumer Reviews and Testimonials, 16 CFR Part 465).
• Start here: the 90-day plan at the end puts your brand entity, your location layer, and honest measurement in order, in that sequence.

What has actually changed for a local brand?

Something real did change, and it is worth naming precisely so you can tell it apart from the hype. On Google, an AI Overview now sits above the results for many searches, and AI Mode can hold a whole back-and-forth conversation. Google's separate Gemini app answers from public information across Google Search and Google Maps. ChatGPT search and Perplexity answer questions with links to the sources they used, and Microsoft Copilot does the same, grounded in Bing's results. For a share of the questions your customers ask, the job is shifting from ranking a blue link they click to being the source the answer is built from.

The mechanics are less mysterious than they sound. Google describes AI Mode as using a “'query fan-out' technique, dividing your question into subtopics and searching for each one simultaneously across multiple data sources” (Google Search Help, AI Mode). In plain terms, the system asks several smaller questions on your behalf, gathers pages that answer them, and writes a summary that links back out. None of these systems has built a new local directory or a private list you can buy your way onto. They read the open web. When someone asks for the best physiotherapy clinic near them or a house cleaner in their suburb, Google's and Gemini's answers lean on the same local pages, profiles, and reviews that already drive your Google local visibility, while the broader web systems lean on how clearly and consistently your locations are described and referenced across the web. The retrieval differs by platform. The material it draws on is the presence you build.

The surfaces your network now appears in, or gets left out of

There are several surfaces worth your attention, and each grounds its answers differently, with its own crawler, its own sources, and its own controls. The traditional fundamentals are the shared foundation, but what sits on top of them changes from platform to platform. Knowing where visibility is actually decided keeps you from paying to influence a place that does not work the way the pitch claims.

Sources for the table: Google Search Central, AI features and your website; Google, Gemini Apps and public information from Search and Maps; OpenAI, publishers and developers FAQ; OpenAI, overview of OpenAI crawlers; Perplexity, crawler documentation; Microsoft, how web search works in Microsoft 365 Copilot. One honest limit runs across all of them: none of these companies publishes a ranking method, so anyone who hands you a list of “AI ranking factors” is selling you a number no platform has ever released.

What hasn't changed, and why that is good news

Almost everything that matters. The plumbing is the same plumbing. Content still has to be reachable by a crawler and in the index, genuinely helpful and written for people, and backed by real experience and trust. Google is unusually direct that trust sits at the centre of the whole quality picture: “Of these aspects, trust is most important. The others contribute to trust, but content doesn't necessarily have to demonstrate all of them” ( Google Search Central, creating helpful, people-first content).

What is striking is how much of the “AI optimization” being sold right now Google specifically tells you to skip. In its own guide it says you do not need to “create new machine readable files, AI text files, markup, or Markdown,” that “there's no requirement to break your content into tiny pieces for AI to better understand it,” that “you don't need to write in a specific way just for generative AI search,” and that “structured data isn't required for generative AI search, and there's no special schema.org markup you need to add.” It even warns that “seeking inauthentic 'mentions' across the web isn't as helpful as it might seem” (Google Search Central, generative AI optimization guide).

We agree with all of it, and saying so costs us a product we could otherwise sell you. For a network, the good news is specific. The local fundamentals a serious brand already invests in, real location pages, accurate profiles, honest reviews, provide the source material for local discovery, especially across Google and Gemini, and they help the broader web systems understand your locations at all. You are not starting a new project. You are being rewarded, or exposed, for how well you already run the old one. The general, non-franchise version of this argument is in AI Visibility Without the Hype.

The two entities every network has to get right

The mental model we use with every multi-location client is the piece most “AI” pitches skip. Search does not see your brand as a pile of pages. It tries to understand your brand, and each of your branches, as an entity: a specific, known thing in the world, with a name, a place, and a reputation. A network has two kinds of entity to keep strong at the same time, and they pull in different directions. The brand has to be known nationally. Every location has to win its own local market against operators who know that market well. Get one right and neglect the other and the AI answer goes to someone else.

Your brand entity: one identity, described consistently

Your brand entity is who search understands you to be across the whole web. You strengthen it by being consistent and by describing yourself clearly in one place. Organization structured data on your main site is how you hand search the facts: your name, your website, your logo, and the “sameAs” links to the profiles that legitimately represent you (Google Search Central, organization structured data). This is how you make yourself legible to search: one clear, consistent description of who you are. Google is clear that structured data earns eligibility, not a guaranteed result: “Google does not guarantee that your structured data will show up in search results, even if your page is marked up correctly according to the Rich Results Test” (Google Search Central, structured data guidelines). The value is that a consistent, well-described brand is easy to resolve as one known thing, which is the whole game when a machine is deciding which cleaning company or clinic the person in front of it actually means.

Your location entities: each branch is its own known place

Every location is its own entity, and it needs its own real presence: its own Google Business Profile, its own page on your site, its own hours, its own reviews. LocalBusiness structured data lets you describe each one to Google, including address, hours, and reviews (Google Search Central, local business structured data). This is where networks get themselves into trouble, usually while trying to be efficient: one profile stretched across several areas, or a single page cloned for every city with the name swapped in. The next section is where you build the local layer properly, because every one of these pieces is easy to get wrong at scale.

We saw the weight of the location layer in our work with Merry Maids Canada, a residential cleaning network we supported for over fourteen years before that engagement ended. Across a network of thirty-five-plus locations, the local websites, not one national page, carried the demand: over a five-year window the network generated more than 1,350,000 genuine leads, with 93 percent coming through the local websites and 83 percent arriving through organic search. Those are historical figures from a completed engagement, and they describe the scale that a properly built local network produced, not a promise of the same numbers for anyone else. The lesson that travels is the structural one. The brand was known nationally, and the locations did the winning.

The local layer AI reads from: pages, profiles, and reviews

If the two entities are the model, this is where you build the location one in practice, across three things: the pages, the profiles, and the reviews.

Local pages that are real, not thin copies

Each location needs a page that is actually about that location: the real service area, the real team, the specifics a local customer would recognize. A find-and-replace template that swaps the city name into the same paragraph is weak for ordinary ranking, and it is weak as grounding for an AI answer, for the same reason. There is nothing there that a person in that market would find unique or useful, and “unique, compelling, and useful” is precisely what Google says influences whether your content shows up in AI search over time (Google Search Central, generative AI optimization guide). A location page is a chance to prove the branch exists and serves that community. Most networks waste it.

One Google Business Profile per location, kept accurate

For local queries, the Business Profile is the engine. Google ranks local results on a combination of “relevance, distance, and prominence,” and it is open about what feeds the last one: “This factor's also based on info like how many websites link to your business and how many reviews you have” ( Google Business Profile Help, improve your local ranking). Two rules keep a network safe here, and Google states both plainly: “Do not create more than one page for each location of your business,” and “including unnecessary information in your business name isn't permitted, and could result in the suspension of your Business Profile” (Google Business Profile Help, guidelines for representing your business). Duplicate listings to cover more ground and names stuffed with services and cities are how a location gets suspended, and a suspended profile has vanished from the exact answers you were trying to win. There is also no shortcut to buy your way up, since Google states there is “no way to request or pay for a better local ranking.” Keeping every profile accurate, at every location, is unglamorous work that quietly decides whether the branch appears when it counts.

Reviews: one of the easiest local signals to get wrong

Reviews feed prominence, so they matter, which is exactly why so many networks reach for the shortcuts that are now illegal or deceptive. Google's own policy prohibits reviews “that have been paid for,” reviews written under a conflict of interest, offering “incentives ... in exchange for posting any review” (the policy's own examples are payment, discounts, and free goods or services), and even “merchants requesting that staff solicit a certain number of reviews” (Google, prohibited and restricted content).

The law has caught up hard. In the United States, a Federal Trade Commission rule in effect since October 21, 2024 bans fake and “AI-generated fake reviews,” paying for reviews of a particular sentiment, undisclosed insider reviews, and using threats to suppress negative ones (U.S. FTC, final rule announcement). That rule is narrower than it first sounds: it “does not prohibit giving incentives for reviews, as long as there isn't an express or implied requirement that the reviews have to express a particular sentiment” (U.S. FTC, reviews and testimonials rule Q&A), though disclosure and other rules can still apply. Google is the stricter one here, prohibiting incentives for any review at all. In Canada, the Competition Bureau treats fake reviews as astroturfing, “the practice of creating commercial representations that masquerade as the authentic experiences and opinions of impartial consumers” (Competition Bureau of Canada, deceptive marketing practices digest).

What you can do is simple and it works: ask every genuine customer for an honest review, at every location, without incentives, without telling them what to say, and without quietly asking only the happy ones. The table below is the line between a review program and a liability.

Sources: Google, prohibited and restricted content; U.S. FTC, Consumer Reviews and Testimonials Rule Q&A; Competition Bureau of Canada. The Google and United States columns reflect published policy and a specific rule. The Canada column reflects the general framework: fake reviews, astroturfing, and undisclosed material connections can be deceptive under the general-impression test, with the outcome depending on the facts, so confirm any Canadian review program with counsel.

Doing this properly pays off, and we have watched it happen. With Comfort Keepers Canada, an in-home senior care brand that operates through locally owned franchise offices, an ordinary, un-gimmicky search foundation is what puts the network in front of families searching for care: in Lime's published results for the brand, it ranks for roughly 7,200 organic keywords, with 156 in the top three positions and 237 search-results features earned. Those are point-in-time figures from a current engagement, not a permanent state, and each office runs its own local profile. No trick produced it. Relevant pages, accurate profiles, and genuine reviews did.

The content that actually earns a citation

When an AI answer cites a source, it is usually citing a page that earned that attention the ordinary way. Google's advice for showing up in its own AI features is not a new checklist; it says “it can be helpful to have a unique viewpoint that stands out,” alongside content people find genuinely useful (Google Search Central, generative AI optimization guide). For a network, that means answering the real questions your local customers ask, from real experience, under a named author, rather than publishing a hundred near-identical keyword pages. Google now asks content to make its authorship and its use of automation evident to readers, posing the test question, “Is the use of automation, including AI-generation, self-evident to visitors through disclosures or in other ways?” (Google Search Central, creating helpful, people-first content). If your network operates in care, health, finance, or law, the bar is higher still, because Google holds content that can affect a person's “health, financial stability, or safety” to a stricter standard. Trust is the through-line, and it comes from content that people who do the work actually wrote.

Is being cited the same as ranking? Mentions, citations, and recommendations

It helps to separate three things that get blurred together in the sales pitch. Being mentioned is your name appearing somewhere on the web. Being cited is an AI answer using your page as a source and linking to it. Being recommended is the answer putting you forward as the right choice for a specific person with a specific need. They are not the same. On Google they are not even won separately, because AI Overviews and AI Mode draw on the same ranking and quality systems as regular results. On ChatGPT, Perplexity, and Copilot the retrieval is their own, but the qualities that earn a citation, relevant and credible pages, well sourced, that a crawler can actually read, are the same ones that earn a ranking. Either way you earn it much the same way, and you measure it differently.

The honest news is that measurement is finally arriving, unevenly. Google's Search Console now has a generative AI performance report for Search, currently rolling out to a subset of website owners. Where you have it, it reports impressions from AI Overviews and AI Mode, “how many times links to your site were shown to a user in a generative AI feature,” broken down by page, country, date, and device, but not the exact queries people typed (Google Search Console Help, generative AI performance report). Microsoft's AI Performance report in Bing Webmaster Tools is more explicit about citations, showing how often your content is cited across its AI surfaces (Microsoft, AI Performance in Bing Webmaster Tools). ChatGPT gives you a more direct read on traffic: it “automatically includes the UTM parameter utm_source=chatgpt.com in referral URLs,” so real click-throughs from ChatGPT show up in your analytics (OpenAI, publishers and developers FAQ). Perplexity and Gemini give you far less. Use what exists, and hold it loosely, because the gap between what you can measure and what actually happened is exactly where a vendor will sell you certainty they cannot have.

What to stop paying for: AI tactics that waste money or create risk

Some of the most confident pitches in this market are for things that do nothing, or things that quietly hurt you. A few to refuse.

The machine-first gimmicks. The special AI files, the content chopped into fragments, the pages rewritten for machines, the campaigns to manufacture brand mentions. Google's own guidance, quoted earlier, says these do not help you in Google Search, and no platform has shown they earn a universal advantage in AI answers, so the main thing they reliably cost is the fee. If a vendor is selling any of them, that is your answer.

The review shortcuts. Everything in the table above. Incentives, staff reviews, bought reviews, gating, and anything AI-fabricated are a regulatory exposure that grows with every location you run.

Blocking the wrong crawler. This one catches careful people. Google-Extended governs whether your content trains future Gemini models and “does not impact a site's inclusion in Google Search nor is it used as a ranking signal” (Google, common crawlers). A separate Search Console control governs whether you appear in AI Overviews and AI Mode, and it too is ranking-neutral (Google Search Console Help, search generative AI control). Over on OpenAI's side, GPTBot is the training crawler, while OAI-SearchBot is the one that makes you eligible for ChatGPT search, and “sites that are opted out of OAI-SearchBot will not be shown in ChatGPT search answers” (OpenAI, overview of OpenAI crawlers). Perplexity runs the same split: its PerplexityBot is “designed to surface and link websites in search results on Perplexity” and “is not used to crawl content for AI foundation models” (Perplexity, crawler documentation). A network that blanket-blocks “AI bots” to protect its content can accidentally erase its locations from ChatGPT and Perplexity while thinking it only opted out of training. Decide training and visibility separately, and check what every one of your properties is actually blocking.

And one piece of reality to keep straight, because it is new. You cannot buy an organic citation. But ChatGPT is no longer ad-free; OpenAI began testing labeled ads inside ChatGPT in the United States on February 9, 2026, initially for eligible Free and Go users, stating that “ads do not influence ChatGPT's answers” and that they run on “separate systems from our chat model” (OpenAI, ads in ChatGPT). Do not let anyone blur the two and sell you “guaranteed AI placement.” Paid ads and earned citations are different things, and only one of them is for sale.

A 90-day plan for a multi-location brand

You do not need a new department for any of this. You need to put the fundamentals in order across the network, in sequence, starting with the foundation and the things closest to a booked customer. This is the plan we run.

Days 1 to 30, get the entities and the profiles clean. Confirm your brand is described in one consistent way everywhere it appears, and add or correct Organization structured data on your main site. Then audit every location's Google Business Profile for the things that get profiles suspended or buried: duplicates, keyword-stuffed names, wrong categories, wrong hours, missing or inconsistent address details.

Days 31 to 60, build the location layer properly. Give every location a genuine local page, not a template clone, and add LocalBusiness structured data to each. Stand up a review routine that asks every real customer for an honest review, the same compliant way at every location, with no incentives and no gating.

Days 61 to 90, earn citations and start measuring. Publish content that answers the actual questions your local customers ask, from real expertise, under a named author. Check whether your Search Console property has the generative AI performance report, and use Bing Webmaster Tools' AI Performance reporting, so you have some read on AI visibility. Finally, audit your crawler controls so no location is accidentally hidden from Google, ChatGPT, or Perplexity.

None of this tells you why one of your locations already out-produces another on ordinary demand and follow-up. That diagnosis is in why some franchise locations generate more leads than others, and the reporting that makes the gap visible is in what a franchise marketing dashboard should show at the location level.

Book a free network growth review

If you would rather find out where your own network is leaking visibility before you spend on any of this, that is what a Growth Opportunity Review is for. We look at your brand entity, every location's profile and page, your review position, and what you can actually measure across the network, and we come back with a ranked list of what is costing you the most and what it would take to fix. You keep the findings whether or not you work with us.

Frequently asked questions

Do we need a separate AI strategy, or one of those “GEO” packages?

Not a separate one, no. The fundamentals that earn you rankings, crawlable pages, helpful content, accurate profiles, and honest reviews, are the same ones that earn AI citations. On Google that is explicit, because AI Overviews and AI Mode run on its core Search systems. On ChatGPT, Perplexity, and Gemini the retrieval is their own, but they still reward the same clear, credible, accessible content. What does change by platform is the plumbing: which crawler to allow, what grounds each answer, and what you can measure. If a package promises to optimize you for AI in general, ask which of its tasks Google has not already told you are unnecessary, and which platform each remaining task is actually for. Usually the list gets very short.

Should we block AI crawlers to protect our brand's content?

Decide it deliberately, and per property, because the controls do different jobs. The trap for a network is the blanket “block all AI” reflex: blocking OpenAI's OAI-SearchBot takes your locations out of ChatGPT's search answers, when all you may have meant to do was keep your content out of model training, which is a separate control. Sort out that distinction for the brand site and each location, rather than reaching for one switch.

Head office runs the brand site and franchisees run the local pages. Who owns AI visibility?

Both, and the handoff between them is where it usually breaks. Head office owns the brand entity: the consistent identity, the main-site structured data, the national reputation. Franchisees, or head office on their behalf, own the location entities: the individual profiles, the local pages, the reviews. Visibility fails when nobody owns the seam, when a location's profile is duplicated or abandoned, or when local pages are published centrally with no local substance. Name an owner for each layer before you spend on anything else.

Can a multi-location brand just get more reviews to win the AI answer?

Genuine reviews help, because reviews feed local prominence. Volume manufactured through incentives, staff posting, or gating does not, and now it carries legal risk in both the United States and Canada. A hundred real reviews earned honestly at a location will beat a review drive built on shortcuts that a regulator or Google can unwind, and it exposes you to nothing. What matters is that the reviews are genuine and that you can stand behind every one.

How do we know if AI search is actually sending us anything?

Better than you could a year ago, though only in patches. Where your Search Console property has the generative AI performance report, it shows your impressions in Google's AI features by page, country, and device, and Bing Webmaster Tools reports citations across Microsoft's AI surfaces. Those measure visibility: whether you were shown or cited. Traffic is a separate question, and ChatGPT tags its referral links with utm_source=chatgpt.com so real click-throughs show up in your analytics. Read what exists per location where you can, and treat all of it as directional, because no tool sees the whole picture yet.

References

1. Google Search Central. Google's Guide to Optimizing for Generative AI Features on Google Search (updated 2026), and AI Features and Your Website (updated 2025).
2. Google Search Central. Creating Helpful, Reliable, People-First Content (updated 2025).
3. Google Search Central. Google's Common Crawlers (Google-Extended), updated 2026.
4. Google Search Central. Organization (Organization) Structured Data (2026); Local Business (LocalBusiness) Structured Data (2025); General Structured Data Guidelines (2026).
5. Google Search Help. AI Mode in Google Search.
6. Google Search Console Help. Generative AI Performance Report (Search), 2026; Search Generative AI Control, 2026.
7. Google. Gemini Apps and Public Information from Google Search and Google Maps. Gemini Apps Help, 2026.
8. Google Business Profile Help. Improve Your Local Ranking on Google, and Guidelines for Representing Your Business on Google.
9. Google. Prohibited & Restricted Content (Contributed Content Policy).
10. OpenAI. Publishers and Developers FAQ; Overview of OpenAI Crawlers; Ads in ChatGPT (2026).
11. Perplexity. Perplexity Crawlers: PerplexityBot and Perplexity-User. Developer documentation, 2026.
12. Microsoft. How Web Search Works in Microsoft 365 Copilot Chat and Agents (2026); Introducing AI Performance in Bing Webmaster Tools (Public Preview), 2026.
13. U.S. Federal Trade Commission. Final Rule Banning Fake Reviews and Testimonials (August 2024); The Consumer Reviews and Testimonials Rule: Questions and Answers; Rule on the Use of Consumer Reviews and Testimonials, 16 CFR Part 465.
14. Competition Bureau of Canada. Deceptive Marketing Practices Digest, Volume 3; False or Misleading Representations and Deceptive Marketing Practices.