A clinic group with four sites and a dozen treatments is told, usually by a keyword tool, that it should have forty-eight pages. Treatment plus town, for every combination. The pages are generated from one template, each with the place name substituted in, and the result is a set of documents that a duplication check will collapse without difficulty.
This is not a marginal problem. It is the leading cause of the crawled, currently not indexed status on clinic sites, and it wastes real budget because the pages have to be produced before they can be ignored.
Why the matrix fails
Two pages are duplicates when their substantive content is the same. Substituting a place name into nine hundred identical words does not make substantive content different, and the system evaluating the page is designed specifically to detect that. What follows is either exclusion or canonicalisation into one arbitrary member of the set, and in both cases the effort produced one indexed page rather than twelve.
There is a second failure that is less discussed. Even where such pages are indexed, they compete with each other for the same queries, so your own site fragments its own signals across a dozen near-identical documents rather than concentrating them in one.
The structure that works instead
Separate the two things the matrix conflates.
- Treatment pages answer a clinical question. They are written once, thoroughly, for the whole group, and they are the pages that compete on informational and treatment queries.
- Location pages answer a logistical question: where is this site, who works there, what is available at this address, how do I get there and park, what are the hours. They compete on place-qualified queries and they support local eligibility.
The location page links to the treatments offered at that site. The treatment page links to the locations that offer it. The cross-product exists as a link graph rather than as a set of documents, which is where it belongs.
| Approach | Pages produced | Typical outcome |
|---|---|---|
| Treatment x location matrix | Treatments multiplied by sites | Mass duplication, most excluded |
| Treatments plus locations, cross-linked | Treatments added to sites | Each page distinct and indexable |
What genuinely belongs on a location page
A location page earns its place when it contains information that exists nowhere else. That means the exact address and how to find the entrance, parking and public transport specifics, the named practitioners who work at that site and their registered qualifications, the treatments actually available there, the hours including any that differ from the group standard, accessibility information, and the direct phone number for that site if one exists.
All of that is verifiable, none of it can be invented, and collectively it is more than enough to make the page substantively different from every other location page in the group. If a location page cannot be filled with real specifics, that is a signal the location may not warrant its own page.
Directories and citations
HIGH CONTROL- The name, address and phone number you submit, character for character
- Which directories you claim and which you deliberately ignore
- The website URL and category on every claimed listing
- Whether an aggregator republishes stale data it scraped years ago
- Whether a directory ranks for your brand name
- The design and quality of the directory itself
- Search the exact phone number in quotes and read every result
- Search the brand name plus the town and audit the first three pages
- Keep one canonical NAP string in a document and diff every listing against it
Consistency across the group
Each site needs its own Business Profile, its own consistent name, address and phone number, and a website URL pointing at that site's own location page rather than at the group homepage. That last detail is missed constantly and it matters, because it is the connection between the profile and the page that describes the same physical thing.
Name consistency is where groups create their own problems. If the profile says "Group Name Manchester" and the location page heading says "Our Manchester Clinic" and a directory listing says "Group Name Ltd, Manchester Branch", three systems now have three strings to reconcile. Pick one form per site and use it everywhere.
Regulated services and public registers
Where a service is regulated, the public register is a durable external record that names locations. In England the Care Quality Commission publishes registration details including registered addresses. Where a clinic is registered, that record is an independent statement of location and provider identity, and it is worth ensuring the name and address you publish match it. Consistency with a public register is an unusually strong entity signal because it is not something a business can quietly change.
Service areas and the pages that should not exist
Clinics that travel to patients, or that draw from a wide area, are tempted to build pages for every town they serve. Pages for places where you have no address and nothing distinctive to say are the thinnest content on the web, and they fail on both counts: they do not qualify you for local results in those places, and they do not earn indexing on their own merits.
If catchment matters, say so on the location page for the site that serves the area, name the areas truthfully in a sentence, and stop there. One accurate paragraph beats twenty pages that exist to contain a place name.
Your own website
HIGH CONTROL- Every byte a crawler receives, and the speed it arrives at
- The entity claims: name, address, identifiers, sameAs, service list
- Canonical URLs, hreflang if used, and the internal link graph
- Whether the page answers the question it was built to answer
- How the answer is displayed once it leaves your server
- Whether a third party republishes an outdated version of your facts
- Crawl the whole site and compare the URL list against the sitemap
- Validate every JSON-LD block against the Rich Results Test
- Measure field data, not only lab scores, for Core Web Vitals
Consolidating a matrix you already have
- Export every URL and group them by treatment and by location.
- For each group, identify the single page with the most impressions and the most substantive content. That becomes the survivor.
- Move any genuinely unique sentence from the others into the survivor. There will be fewer than you expect.
- Redirect the rest permanently into the survivor, and update every internal link to point at the final destination.
- Rebuild the location pages with real local specifics, and cross-link them to the treatment pages.
- Resubmit the sitemap and watch the non-indexed count in the Pages report fall over the following weeks.
What to expect afterwards
Total indexed pages will fall, sometimes sharply, and that is the intended result. The measure that matters is impressions and clicks to the surviving pages, and whether the pages you meant to rank are now the ones receiving queries. A smaller, denser site is easier to crawl, easier to keep accurate, and considerably easier to write well, which is the constraint that was binding all along.