Ten location pages that share one name, one service list and one paragraph of copy give an engine no way to tell a group from a practice with ten offices. It merges them, splits them, or credits every answer to whichever page it saw first. The fix is structural: give every location its own identity, page and proof.
A location page and a service area page are different jobs. A location page is for an office you can walk into, and a service area page is for a neighbourhood you serve. Service area pages covers the second one.
- /locationsAll addresses in one list
- /locations/northIts own address, phone and hours
- /locations/downtownIts own address, phone and hours
- /locations/southIts own address, phone and hours
The Ambiguity Problem
A single practice with one office has one name, one address, one phone number and one set of hours. A group has the same group name repeated across ten addresses, with a phone number that changes, a provider list that changes, and hours that change by week. If those two things look the same in your data, the engine has to guess which one you are, and it will not ask.
The guess goes wrong in predictable ways. Either the locations collapse into one entity, so a patient in your third city is sent to your flagship and the answer names one office, or they separate so far that the group brand never reaches them and every office competes with every other for the same terms. Neither is a copywriting problem. Both are identity problems.
The Fingerprints
Location pages with identical titles. One phone number on all ten pages. The same paragraph about the group copied into every footer. A single reviews block with no location attribution. If your site looks like this, the engine's conclusion is reasonable and you are the one who has to change it.
Give Each Location Its Own Identity
Each location needs its own identity across seven fields. Some of them are administrative and some are editorial, but leaving one shared undoes the other six.
Swipe the table sideways to see every column.
| Field | What It Must Be At This Location | Why It Cannot Be Shared |
|---|---|---|
| Name | The group name plus this location, in the form patients say it | Two locations with one name cannot both be right in one answer |
| Address | Full street address with the suite, formatted the same way everywhere | A shared or missing suite splits the entity |
| Phone | A line that reaches this office rather than the central desk | A number answered at another office is not local |
| Hours | The days this office opens, with the exceptions entered | Local questions are opening-time questions |
| Providers | The dentists and hygienists who work here, each with their own page | A shared roster makes the offices interchangeable |
| Services | The treatments delivered here, not the whole group menu | Claiming a service you do not offer costs trust |
| Reviews | The reviews left for this office, shown on this office's page | Group-level reviews belong to nobody in particular |
One Shared Phone Number Is The First Thing To Fix
If every location rings through to the same central desk, the strongest local signal a location page can have is missing from all of them. A dedicated line per office is the fix, and it is worth doing even when the central desk is more convenient, because the number is also part of the identity an engine matches on.
The Location Page Anatomy
Ten sections, each doing a job a patient or an engine can use:
- A unique title and H1 naming the location and the city, not the group name repeated
- An opening paragraph written for this neighbourhood: who is treated here and what is different about this office
- The address block with the suite, a map and how to find the entrance
- Hours for this office, including the days it is closed and the holiday exceptions
- The phone number that reaches this office and a booking link that opens this location
- The provider list for this office, each provider linked to their own page
- The services this office delivers, described here rather than copied from the group hub
- The reviews left for this location, with the location named in the source
- The questions patients actually ask about this office, answered on the page
- Links to the group hub and to the sibling location nearest to the reader
Stopping Cannibalisation
Seven rules, applied in this order:
- Give every location its own title and H1. Two pages that open with the same words are two candidates for one intent, and the engine will pick one or treat both as duplicates.
- Make each page canonical to itself. A location page should point at itself, never at the group hub and never at a sister office.
- Stop copying service copy. A treatment page identical at ten locations is a duplicate; describe the treatment once at group level and let the location pages carry the local facts.
- Split the intent. Decide which city and which questions each location answers. Two offices serving one city should not target the same phrase.
- Link like a hub and a spoke. The hub links to every location with anchor text naming the office and the city; each location links back to the hub and to its nearest sibling rather than to all ten.
- Do not noindex a thin location. A thin page that competes is a problem, but a hidden page cannot resolve. Fix the page or merge the location.
- Keep the naming pattern constant. The same name in the title, in the profile, in the review request and in the structured data, every time.
Structured Data For a Group
Two levels of markup. The group page carries Organization markup with the locations listed; each location page carries LocalBusiness markup, or the the Dentist schema type, with its own address, telephone, opening hours, coordinates and a parent organisation pointing back to the group. Google's guidance on AI features asks for the marked-up facts to match what a person sees on the page and warns that duplicated pages waste crawl effort, which is exactly what ten identical location pages do.
Do Not Mark Up The Same Address Twice
If two pages carry markup for the same address, the engine gets two records for one office. Mark each address up once, on the page that represents it, and let the group markup live on the hub with the list of locations. The guide to dental schema markup covers the properties and the order they go in.
What Multi-location Changes In Your Measurement
A group cannot be measured with one question set. The audit runs forty questions across four engines, and for a group you split that: one set per location for local intent, one set for the group brand, and one set for the treatments you want quoted. drsmiles.ai reports them separately, because a group score averaging ten locations tells you nothing about the office that needs fixing.
Ask each location a different thing. Best dentist in Eastside belongs to one office. Family dentist accepting new patients belongs to the group. How much do clear aligners cost belongs to no location page at all: it belongs to a treatment page, and the locations should link to it rather than compete with it. The guide to service area pages covers where the line between a location and an area belongs.
Then watch for cannibalisation in the results. If two locations get named for the same question, the pages are still too similar, and the fix is in the pages rather than in the measurement. Answer sets vary by wording, model, location and time, so run the same set several times before you call a change a win.
If the location pages do not exist yet, or they exist and all read the same, that is a build problem before it is a content problem, and our Smart AI Websites build is where it starts. The facts still have to agree across every listing, which is the same name, address and phone cleanup as any other practice, multiplied by ten. The case studies show how the measurement reads afterwards, and the blog index covers the rest.