usdigitalmarketingindex.com SEO
How can near-duplicate content be identified across hundreds of template-based location pages?
Updated November 15, 2026
Read a sample of the pages side by side, mark which text is shared template and which is specific to each place, and then record for every page what location-specific material it contains. A page is a near-duplicate when the only things that change from one location to the next are the place name and a few inserted values. Similar wording is not the fault in itself. Pages for a chain’s branches will share structure and policy text. The question for each page is whether a visitor in that place gets anything from it that the page for the next town would not give them.
Finding repeated textual blocks
Start by separating the template from the content.
- Choose a sample across the set: large and small locations, old and new pages, different regions.
- Open three or four in adjacent windows and read them in parallel, section by section.
- Mark each block as one of three kinds: identical on every page, identical apart from the place name, or written for this location.
- Record the template’s structure: which sections exist, and which of them ever contain location-written text.
A manual comparison of one section from three pages shows what to look for. The excerpts are invented.
| Page | Excerpt from the "About our service" section (invented) | Classification |
|---|---|---|
| Springfield | Our Springfield team provides fast, reliable drain cleaning to homes and businesses across Springfield. | Template with name swapped |
| Riverton | Our Riverton team provides fast, reliable drain cleaning to homes and businesses across Riverton. | Template with name swapped |
| Lakewood | Our Lakewood crew works from the depot on Mill Road. Many Lakewood homes built before 1970 have clay sewer laterals, so we carry root-cutting heads on every truck here. | Written for this location |
The first two would read identically with the names removed. The third would not.
Crawling tools can report textual similarity between pages, and that helps find candidates in a set of hundreds. Treat the score as a pointer to pages worth reading. A high score does not prove a page is useless, and a low score achieved by rewording the same sentences does not prove it is useful.
Assessing user-facing differences
For each page, ask what a local visitor can do or learn there.
- Is there an address, opening hours, and a phone number that belong to this location?
- Are the services listed the ones offered here, where branches differ?
- Are prices, availability, or terms specific to this place?
- Are there directions, parking details, or a service-area description that only apply here?
- Are there staff, photos, or reviews from this location?
Functional differences count even when the prose is shared. A branch page with its own hours, stock, and booking calendar is useful to someone looking for that branch, whatever its introductory paragraph says.
Identifying pages with no local proof
Record the findings in a matrix, one row per page. The rows below are sample rows.
| Location page | Own address and hours | Local service details | Local proof: photos, reviews, staff | Local conditions or rules | Unique text beyond the name | Finding |
|---|---|---|---|---|---|---|
| Springfield | Yes | No | No | No | No | Functional only |
| Riverton | No, served from Springfield | No | No | No | No | No local substance |
| Lakewood | Yes | Yes | Yes | Yes | Yes | Substantive |
| Hillcrest | No | No | No | No | No | No local substance |
Three groups emerge.
- Substantive: the page has real local content. Leave it.
- Functional only: the page has a real location behind it and little else. It serves people seeking that branch, and it can be improved.
- No local substance: no premises, no local details, no proof. The page exists to carry the place name.
The third group is where the risk lies. Google’s spam policies give, as an example of doorway abuse, “Having multiple domain names or pages targeted at specific regions or cities that funnel users to one page” (Google’s spam policies for web search). Pages in that group that all lead to one contact form match the description closely. Whether a given set is treated as a violation is Google’s judgment. What the matrix establishes is which pages offer a local visitor nothing.
Editorial remediation priorities
Decide per group, in order of risk.
| Group | Evidence | Outcome |
|---|---|---|
| No local substance, and no real operation in the place | Matrix row empty | Remove the page and redirect it to the nearest real location or a service-area page that names the town |
| No local substance, and the place is served | Matrix row empty, service confirmed | Fold into a shared service-area page, or gather real local material and rebuild |
| Functional only | Address and hours present | Add what is true and local: services at this branch, staff, photos, conditions, reviews |
| Substantive | Several columns filled | Keep; use as the model |
Sequence the work.
- Deal with the empty pages first, since they carry the most risk and the least value.
- Collect local material for functional pages from the people who work there. It cannot be written from head office.
- Change the template so that location-written sections are required fields, and a page cannot be published with them blank.
- Recheck the set after changes with the same matrix.
Do not solve the problem by rewording. Running the same paragraph through variations for each town lowers a similarity score and leaves every page as empty as before.
Whether neighboring towns should have separate pages at all, or share one, is a planning decision made before pages are built. That decision is covered in the guide to when nearby cities should share a single service-area page.
The test for any single page: delete the place name, and see whether anything on the page still tells you where it is about.
Add your comment to this answer