AI search for multi-location businesses: keep each location's hours, address and services from being mixed up
When a business has several locations, most wrong AI answers come from one location's facts being used for another. Give each location its own page with its facts in plain text (address, phone, hours, temporary closures, what differs there), tie it to the brand in structured data, and keep the same values in Google Business Profile and your other listings. Those three steps are the core of the work.
Written November 15, 2026 (JST). This guide is for any business with two or more locations or offices (restaurants, salons, clinics, shops, schools, estate agents) and for the agencies that build their sites. The company itself (legal name, founding date, leadership) is covered in About page AI search optimization, and the basics of one location's markup in local business schema and SEO plugins. This guide is about the problems that come from having many locations.
AI answers change by day and by wording. These steps make the information AI reads accurate; they do not guarantee that you will appear in an answer or in any position.
How AI answers go wrong for multi-location businesses
ChatGPT search, Perplexity and Google's AI Overviews read web pages, build an answer and show their sources. With one location, "the business's facts" are a single set. With several, the pages an AI reads contain several addresses, several sets of hours and several phone numbers. If the text does not make clear which value belongs to which location, the answer can be built from the wrong one.
The questions people ask split into questions about the brand and questions about one location:
- "Is the Umeda branch of [brand] open on Sundays?" (that location's hours and closing days)
- "Which [brand] locations have parking?" (facilities that differ by location)
- "Is the Shinjuku [brand] still open, or did it move?" (closures and moves)
- "Is there a [brand] near Kyoto Station?" (which location serves which area)
- "Can I get the same menu / treatment / service at every [brand]?" (services that differ by location)
- "Who runs [brand]?" (brand and operating company)
Four mistakes come up again and again: the flagship's hours given as every location's hours, a closed location described as open, the central booking number given as a location's phone, and a service only some locations offer described as available everywhere. Each becomes less likely once every location's facts are written on that location's own page.
Google says there are no additional requirements and no special structured data for appearing in AI Overviews or AI Mode, and that the usual SEO basics apply (Google Search Central: AI features and your website). Multi-location work is the same: not a special way of writing for AI, but each location's facts published accurately, readably and with the same values everywhere.
Site structure: separate the brand page, the location list and the location pages
First decide which page answers which question. A single page that lists every location makes it easy for an AI to attach a paragraph to the wrong location.
| Page | Questions it answers | What to write | Structured data |
|---|---|---|---|
| Brand (home or About) | "What is [brand]?" "Who runs it?" | Brand and operating company, services common to all locations, number of locations (with a date) | Organization (head office) |
| Location list | "How many [brand] locations in Osaka?" "Is there one near Kyoto Station?" | Names and addresses by region and city, linking to each location page | Optional (put it on each location page) |
| Location page (one per location) | "Is Umeda open on Sundays?" "Is there parking?" | Facts that belong to that location only (next table) | One specific LocalBusiness type per location |
| Closure or move notice | "Is Shinjuku still open?" | Closing date, new address, links to nearby locations | Remove the location's structured data |
Give each location a URL that will not change, such as /stores/osaka/umeda/. Renaming locations or URLs midway creates more mismatches with what AIs and other sites read before. If you must change one, redirect the old URL to the new one.
Do not create pages for areas where you have no location (pages that only swap the place name, "looking for X in [town]? Choose us"). They can also lead an AI to describe a location that does not exist.
What each location page needs, and the matching properties
The table lists what to write in plain text on each location page and the matching schema.org property. Google's local business documentation lists name and address as required, with openingHoursSpecification, telephone, url, geo and department among the recommended properties, and recommends the most specific type available, such as Dentist, HairSalon or Restaurant (Google: Local business structured data).
| Item | How to write it | Typical AI mistake | Property |
|---|---|---|---|
| Location name | A fixed pattern, brand + place (e.g. Sample Dental Umeda), the same for every location | Mixed with another location, known only by a nickname | name |
| Address | Postcode to building and floor, plus the nearest station and exit in a sentence | A previous address, or a nearby location's address | address, geo, hasMap |
| Phone | Say whether it is the location's direct line or a central booking line | The central number given as the location's | telephone |
| Hours and closing days | Per location, on each location's page, even if they are the same everywhere | The flagship's hours given for every location | openingHoursSpecification |
| Temporary closures and holidays | With dates; remove them when they are over | Old closure notices left in place | specialOpeningHoursSpecification |
| Services and facilities | Separate what every location offers from what this one offers ("This location does not offer Saturday orthodontic consultations") | One location's service described as available everywhere | Body text, FAQ, department (a section inside the location) |
| Parking, payment, languages | Spaces or partner car parks, payment methods, languages spoken | Another location's facilities | Body text (paymentAccepted and others) |
| Store code | The number used internally and in Google Business Profile | Two locations with similar names confused | branchCode |
| Link to the brand | "Operated by Sample Co., Ltd." with a link to the About page | Franchisee and franchisor confused | parentOrganization |
| Opening date, last updated | "Updated on [date]" at the end of the page | No way to tell how current the facts are | Body text |
Location pages often end up as the same text with only the name and address swapped. That is acceptable, but writing at least three facts that only apply to that location (facilities, which services it handles, busy times, directions) gives an AI something to tell the locations apart by.
Keep each location's FAQ on that location's page, separate from brand-wide questions: questions that can be answered with the location's name in them, such as "Is there parking at this location?" or "Can I get X at this location?"
Closures, moves and temporary closures
The wrong fact that lasts longest for multi-location businesses is a closed location. Deleting its page does not remove what other sites and AIs have already read. Clean up in this order:
- Do not delete the closed location's page straight away. Rewrite it as "This location closed on [date]" with links to nearby locations. For a move, redirect to the new location's page or state the new address clearly.
- Remove it from the location list and remove its structured data (do not keep publishing a closed location as a LocalBusiness).
- Update its Google Business Profile to show the closure or move.
- Search the rest of the site (footer, recruitment page, campaign pages, PDF flyers) for mentions.
- Ask other publishers (booking sites, portals, local directories) to update their listings.
- For a temporary closure, give dates and show it with specialOpeningHoursSpecification or a notice. Remove it when you reopen.
Structured data: tie each location to the brand
For several locations, structured data should make "which company, which location" readable by machines. The approach:
- Publish the head office (brand or operating company) once as an Organization on the home page or About page, with a fixed @id such as https://example.com/#organization.
- On each location page, publish exactly one LocalBusiness of the most specific type, with an @id unique to that location such as https://example.com/stores/osaka/umeda/#store.
- Point from each location to the head office with parentOrganization. schema.org marks the older branchOf as deprecated and recommends parentOrganization instead.
- Put the store code in branchCode, which schema.org itself describes as the "store code".
- If a location contains a separate section (schema.org's examples are a store with a pharmacy and a bakery with a cafe), connect it with department, which lets each section have its own hours.
- Put each location's values on its own page rather than listing every LocalBusiness on the location list. Do not put values in structured data that the page text does not show.
Google lists matching structured data to the visible page content as one of the basics and recommends keeping Business Profile information up to date (AI features and your website). Treat structured data as the answer key for what the location page already says.
Keep Google Business Profile and the site on the same values
When a location's Business Profile and its page disagree, neither people nor AIs can tell which is right. What helps most is one master sheet of every location's facts, from which the site, Google Business Profile and booking sites are all updated.
| Column | What to decide |
|---|---|
| Store code | The same number internally, on the site and in Business Profile |
| Location name | The fixed brand + place pattern |
| Address and phone | One way of writing numbers and separators |
| Hours, closing days, temporary closures | When they change and who updates them |
| Services and facilities | Common to all vs this location only |
| Location page URL | Never changes; redirect if it does |
| Status | Open, temporarily closed, closed (with dates) |
| Last checked | On a fixed cycle, such as monthly |
Bulk verification for businesses with many locations, tools for managing several locations together and the handling of store codes all come with conditions Google sets, and those conditions change. Check the current version in the Google Business Profile Help Center. If location managers edit their own profiles, agree that the master sheet is updated first so the two do not drift apart.
On WordPress (with Nurevo AEO)
On WordPress, Nurevo AEO (free) publishes the head office's facts as structured data and in llms.txt; combine it with your location pages as below. Nurevo's settings describe one business per site, so enter the head office (brand) there.
- Write the location pages first, following the table above (text before structured data). Use one page or one entry of a location post type per location.
- In Nurevo's business details, enter the head office's name, description, address, phone, URL and email. Unlike a single-location business, entering one location's opening hours here makes them look like every location's hours.
- Enter each location's own questions in the Nurevo AEO panel of that page's editor as this page's own questions, and place the Nurevo FAQ block set to "This page's own questions". They are published as FAQPage on that page only.
- To publish a LocalBusiness per location, use the developer filter webmcp_canary_schema_graph to add the location's node to extra on location pages only. The head office's @id is available as $bundle['org_id']; put it in parentOrganization. Build the location's values from the same source as the page text (custom fields, for example).
- With Yoast SEO, Rank Math, All in One SEO or Slim SEO, Nurevo adds only the facts that plugin's structured data lacks, and new nodes (such as a location node) are appended to that plugin's graph. Nurevo also works alongside SEO SIMPLE PACK and VK All in One Expansion Unit (from 0.9.1).
- If something else, such as a store-locator plugin, already publishes structured data per location, avoid duplicates: tick "Publish no Nurevo structured data on this page" on those pages, or choose one source.
- Open /llms.txt and check that the location list and your main location pages are there. llms.txt lists the 30 most recently updated pages and posts, so with many locations some pages will be missing. The webmcp_canary_llms_txt_lines filter can add a line for the location list, and a closed location's page can be removed with "Leave this page out of llms.txt".
- If franchise locations run separate WordPress sites, export the settings with wp nurevo export, import them on the other site with wp nurevo import, then change only the name, address, phone and hours (keys and tokens are not exported).
The same checks apply to sites not built on WordPress. Enter a URL in the free check, no sign-up needed, to see the state of structured data, llms.txt and AI crawler access. You can check location pages one URL at a time.
How to check: per location, with and without the name
Check by picking locations and recording results per location. You do not need to ask about every location every time: start with locations whose hours or address changed, those that closed or moved, and those that get the most enquiries. Ask the same question of the same AI on different days, and do not judge from one answer.
| What to record | Example |
|---|---|
| Date, AI, question | Nov 15, Perplexity, "Is Sample Dental Umeda open on Sundays?" |
| Facts in the answer | Are the hours, address, phone and services this location's, or another's? |
| Sources | This location's page, the location list, or an outside booking site? |
| Question without the name | Does the location appear for "dentist open on Sunday in Umeda"? |
| Closed locations | Is a closed location described as open? |
Nurevo's free AI check (once a month, for connected Free and Standard sites) asks questions built from the business type and area in your settings, without your name. There is one area in the settings, so check your other locations' areas by hand with the table above. Pro's visibility measurement (beta) currently covers Perplexity. The full method is in how to measure AI visibility.
Multi-location checklist
- Every location has its own page, at a URL that will not change.
- Location names follow the same brand + place pattern everywhere.
- Each location page states that location's address, phone (direct or central), hours and closing days in text.
- Temporary closures and holidays have dates and are removed when over.
- Services common to all locations are written separately from those only this location offers.
- Closed or moved locations show a notice and are removed from the location list and structured data.
- Each location's LocalBusiness has its own @id and points to the head office with parentOrganization.
- Structured data values match the location page text.
- The site and Google Business Profile are updated from the same master sheet.
- Locations can be browsed by region, with links to each location page.
- You picked locations and checked questions with and without the name on different days, and recorded the results.
FAQ
Do we need a page for every location, or is one location list enough?
To answer questions about one location (its hours, its facilities), one page per location is more reliable. With only a list, an AI can easily attach a value to the wrong location. Even with few locations, give each one its own heading with its address and hours together.
If every location has the same hours, do we still need them on each page?
We recommend it. An AI may answer from the location page alone, and if the hours are not there it will use values from other pages or other sites.
Should we use branchOf or parentOrganization?
schema.org marks branchOf as deprecated and recommends parentOrganization. Point from each location's LocalBusiness to the head office Organization's @id with parentOrganization.
Should we delete a closed location's page straight away?
Turning it into a notice with the closing date and links to nearby locations, and removing it from the location list and structured data, communicates the closure better than deleting it. For a move, redirect to the new location's page or state the new address.
Our franchisees run their own sites and social accounts. How do we keep them consistent?
Treat the head office's master sheet as the source of truth, share the agreed spelling of names, addresses, phone numbers and hours with franchisees, and ask them to use the same values. Linking both ways, from the franchisee's site to the brand's location page and back, also makes the relationship clear.
We have too many locations for llms.txt. Should all of them be listed?
No. If the location list and your main location pages are there, the rest can be reached from them. In Nurevo AEO, the webmcp_canary_llms_txt_lines filter can add a line for the location list.
Will these steps make every location appear correctly in AI answers?
That cannot be guaranteed; AI answers change by day and wording. These steps reduce the causes of one location's facts being mixed with another's and make the information AI reads accurate and consistent.