What you need before you start
You need a clean inventory first: every real location mapped to one verified Business Profile and one page on your site, all owned by the brand. Errors here repeat later, which is why our franchise SEO service starts with this audit.
Google allows one profile per location and asks chains to keep names and categories consistent, according to its guidelines for representing your business. Extra details in a name, such as "near SOHO", are rejected unless they are part of the real-world name. A business managing 10 or more profiles can request bulk verification; service-area businesses cannot.
Set up the account structure once. A business group works like a shared folder: its owners and managers reach every profile in it. Keep the brand as primary owner and add regional staff or franchisees as managers, who can edit but cannot remove users. Give each location a unique store code; changing it in a spreadsheet upload creates a duplicate.
Your inventory sheet should hold, per location:
- store code and Google place ID
- location page URL
- primary and secondary categories
- owner of record and managers
- status on Apple Business and Bing Places
Cover the other maps too: Apple Business replaced Business Connect on 14 April 2026 and takes bulk files of up to 10,000 locations, and Bing Places relaunched in October 2025 with bulk editing.
The step-by-step playbook
The playbook is six jobs repeated for every location. Each is cheap per branch and expensive to skip across a whole chain.
- A page for each real location
- Profile linked to that page
- Reviews where they are thin
- LocalBusiness markup
- Crawlable store locator
- AI answers, per city
Citemill dataset, September to October 2026. Bases: 475, 174, 325 and 562 chains or pages.
Step 1. Build one page per location that describes a real place
A location page answers what a customer asks before visiting: address, hours, parking, who works there, which services that branch offers, and how to book. Google's spam policies name the failure case: "pages targeted at specific regions or cities that funnel users to one page" count as doorway abuse.
How much unique text a page needs has never been measured. Vendor reports show gains after rewrites, such as a six-location UK group going from 38 to 191 leads a month (Indexed, July 2026), but each bundles several changes. Status: advised, not measured.
Step 2. Point each Business Profile to its own location page
We found that 34% of chains with location pages (109 of 325) send at least one profile to the homepage. Google asks for a website that "represents your individual business location", and the matching location page is the plain reading.
The ranking effect is weaker than guides claim. Ultimate Analytics followed 3,438 Maps listings in October 2026: 78% of those linking to a website pointed to a homepage, and among independents a city page showed no significant ranking advantage.
Ultimate Analytics, October 2026, 3,060 listings with a website, nine trades, 22 US cities.
The traffic effect is clearer. A UK retailer that repointed both profiles and added markup saw shop-page clicks rise from 7 to 358 in matched 28-day windows (Seven Marketing, 2026, vendor-reported). Status: neutral for rank, positive for branch traffic.
Tag each listing link with UTM parameters, such as utm_campaign=gbp-store-0142, so branch traffic shows in analytics.
Step 3. Fix reviews where they are thinnest
Across 475 chains with at least two reviewed listings, we found a median gap of 6x between the most-reviewed and least-reviewed location. In 84% of chains the gap was 2x or more, and in 55% it reached 5x.
Citemill dataset, 475 chains with two or more reviewed listings, October 2026.
Customers judge one branch at a time: in BrightLocal's 2026 survey of 1,002 US adults, 47% avoid businesses with under 20 reviews.
Count matters more than pace. Localo's study of 16,098 profiles (September 2026) put top-3 rates at 39.3% with 1 to 5 reviews and 73.0% at 201 to 500, while review velocity moved the rate only from 70.2% to 73.5%. Status: count measured (correlation); velocity weak.
Localo, September 2026, 16,098 profiles. Correlation, not causation.
Ask at the weakest branches first, at a steady pace: after one QR push doubled monthly volume, Google removed 342 of 595 reviews and paused new ones for 30 days (Sterling Sky, September 2026).
Step 4. Add LocalBusiness structured data to every location page
We found that 56% of the location pages we checked (97 of 174) carry no LocalBusiness markup with that location's address. Google's documentation, last updated 8 September 2026, requires name and address. It recommends the specific location's URL, coordinates with at least five decimal places, phone and opening hours, and supports a nested department for units such as an in-house pharmacy.
Use the most specific type, such as Dentist, one block per page, with values identical to the Business Profile and parentOrganization pointing at the brand. Status: documented by Google, not measured for rank.
Step 5. Make the store locator crawlable
Google follows standard links with an href attribute, so a locator that loads branches by script can hide your pages. Give every location a static URL linked from a hub, city pages and nearby branches. Status: advised, not measured.
Step 6. Check AI search, branch by branch
Access is rarely the issue: only 2 of 562 chains (0.4%) block ChatGPT's or Perplexity's search crawlers. Check anyway: OpenAI says sites blocking OAI-SearchBot "will not be shown in ChatGPT search answers". For AI Overviews, Googlebot is the control.
Inclusion depends on the data AI tools read. In BrightLocal's September 2026 study of about 200,000 localized searches, Google Business Profile made up 28.6% of citations and Yelp 9.5%. A Steady Demand study of 2,880 prompts (August 2026) found Yelp in 95.8% of the data behind ChatGPT's business cards. Claim each location on Google, Yelp, Apple and Bing.
BrightLocal, September 2026.
Spearleaf, August 2026, 216 US markets.
Location matters too, as Spearleaf's August 2026 data above shows, so each branch page and profile should state its city plainly.
On 12 October 2026 we asked ChatGPT, with web search set to the city, one realistic question for each of 108 chains, such as "best dentist in Sydney". The chain was named in 16 answers (15%). Answers vary: a London clinic group replied with a screenshot of its own run, where one of its clinics sat third. Spearleaf saw only 48.2% of named businesses named again days later, so track a fixed prompt set over weeks.
Why a Maps top 3 does not get you into AI answers
Ranking well in Google Maps does not mean ChatGPT or Perplexity will name you. In our benchmark of six cleaning franchises across 10 US cities, one brand sat in the Maps top 3 in three cities and was named in 0 of 18 ChatGPT and Perplexity answers for those cities.
SOCi reported a 0.72 correlation between local performance and AI visibility in November 2025, but a correlation across thousands of brands says little about one branch in one city, and location-level numbers diverge, as the chart below shows:
- Only 11.6% of ChatGPT picks were in the Maps top 3 (Suff Digital, September 2026, 17,464 recommendations in 79 US cities).
- Chains made up 3.9% of ChatGPT picks against 6.3% of the Maps top 20 (Suff Digital, September 2026).
BrightLocal, September 2026; SOCi, January 2026, 2,751 brands.
Methods differ, ours included, so compare directions rather than exact percentages.
Mistakes that cost the most
The expensive mistakes are quiet ones that repeat across every branch: profiles linking to the homepage, pages without markup, branches with a fraction of the brand's reviews, and moves handled as new listings.
| Check | Result | Base |
|---|---|---|
| At least one profile links to the homepage, not the location page | 34% (109 chains) | 325 chains with location pages |
| Location page has no LocalBusiness markup with its address | 56% (97 pages) | 174 location pages checked |
| Median review gap, most-reviewed vs least-reviewed location | 6x | 475 chains with 2+ reviewed listings |
| Blocks ChatGPT or Perplexity search crawlers in robots.txt | 0.4% (2 chains) | 562 chains |
| Chain named by ChatGPT for one local question (12 Oct 2026) | 15% (16 answers) | 108 chains |
Two more mistakes show up often:
- Creating a new profile when a location moves. Google says to update the address, and reviews follow if the name stays the same.
- Marking short closures as "temporarily closed". Google reserves it for closures over seven days and says such profiles can rank after open businesses for broad queries.
How to check it works, location by location
Measure each branch as its own business, with four views: Search Console for its page, Business Profile performance, tagged calls and bookings, and fixed AI prompts for its city.
In Search Console, add a URL-prefix property per location folder or region; each allows up to 100 non-owner users. Clicks from AI Overviews and AI Mode count in the standard Performance report, while the generative AI report, global since 31 August 2026, shows impressions only.
Sort every report by the worst branch and re-run the table checks each quarter. Our sample report shows what a per-location view looks like.
When this playbook doesn't fit
Two situations need a different plan: service-area brands without storefronts, and franchise systems where franchisees control the listings.
Service-area brands get one profile per base and must hide the address, which can cost visibility: Sterling Sky saw rankings and calls fall when an address was hidden and recover when it returned (September 2026). BrightLocal measured AI mentions of 33.6% for storefronts versus 29.4% for service-area businesses.
Franchise systems usually fail on ownership: franchisee-made profiles bring duplicates and conflicting phone numbers. We found no study comparing franchisee-run and corporate-run listings, so this rests on governance: the franchisor as primary owner, franchisees as managers, and a written exit process for resales.
In healthcare, separate practitioner and department profiles can compete with the practice listing: Sterling Sky saw an orthodontist's practice listing improve after a practitioner listing was removed (March 2025, anecdote). Our guide to local SEO for dentists goes deeper on dental groups.
Questions
What is multi-location SEO?
Multi-location SEO is search optimization for a business with several physical sites, done per site rather than per brand. Each location gets its own verified Business Profile, page with markup, reviews and tracking. Success is measured branch by branch, in Maps, organic results and AI answers for its own area.
Should each location have its own page or its own website?
Its own page on the main domain, in most cases. One domain keeps links and authority together, and Google's LocalBusiness examples use per-store URLs under one site. Separate city domains that send visitors to one place match Google's description of doorway abuse. A page per location, linked from a crawlable locator, does the job.
Can one Google Business Profile cover multiple locations?
No. Google's guidelines call for one profile per location, and a location shown with an address needs permanent signage with the business name. Service-area businesses get one profile per base. Departments and practitioners may have their own profiles when they serve the public directly.
How do you verify many Google Business Profiles at once?
Use bulk verification. It opens when one business manages 10 or more profiles, and you request it from the Verifications tab in Business Profile Manager. Agencies managing different businesses and service-area businesses cannot use it. Use an email on the brand's own domain, and note that a spreadsheet upload alone sends no request.
Do location pages count as duplicate content?
Shared templates are fine. Google says some duplicate content on a site is normal; near-identical pages are clustered and one is shown as canonical, so the others fade from results. The real risk is pages built to rank in each city and funnel users to one place, which Google treats as doorway abuse. Give each page facts only that branch has.
How do you get a location named by ChatGPT?
Make sure OAI-SearchBot can crawl the site, then work on the sources ChatGPT reads. Studies from 2026 show heavy use of Yelp and Business Profile data in local answers, so claim each location on Google, Yelp, Apple and Bing. Answers change between runs, so track fixed city prompts over several weeks.
