Why duplication no longer works
The technique used to work: generate thirty « Plumber in [town] » pages from a template, and wait. It no longer works, for a mechanical reason.
Google groups pages whose content is near-identical and keeps only one, deemed canonical. The other twenty-nine are not « penalised » in the sense of a sanction: they are simply ignored. The result is the same, with an added cost: your crawl budget is spent on pages that return nothing, at the expense of the ones that matter.
A nastier effect stacks on top, cannibalisation: thirty pages targeting the same intent compete for the slot. Google picks one, often not the one you would have picked.
What makes a location page useful
The test is simple: remove the town name from the page. If the text stays true word for word for any other town, the page has no content of its own.
What gives it content of its own, in order of ease:
- A local piece of work. A job, an intervention, a customer, with what was done and the particular constraint met. It is the most convincing content and the hardest to copy.
- Genuinely local service information. Response time from your workshop, days you pass through, area covered without travel charges, parking for a shop.
- A constraint of the ground. Planning rules in a conservation area, soil type, ageing pipework in an old district: these are subjects your customers actually type.
- A named contact. The name and face of the person who works that area.
A page holding two of those four has a reason to exist. A page holding none is a duplicate with a different proper noun.
How many to create, and in what order
The reflex is to want to cover the whole service area. That is the mistake: thirty empty pages are worth less than one rich page. The right number is the one you can feed, today, with content of its own.
An order that works for a services business:
- One page per town where you have already done something worth telling. Start there, the content exists, it is just sitting in your phone.
- Then the towns that weigh most in your revenue, even without publishable work: service information is enough to start.
- The rest of the area stays covered by a single page listing the towns served, with no dedicated page. An honest list beats thirty shells.
Two technical details that matter: every location page must have its own title and its own meta description, and be linked from the navigation or a hub page. A location page nothing links to is an orphan, and it will receive almost nothing.
How to link location pages to each other
A well-written but isolated location page is close to useless. That is the point most guides skip: content of its own solves the duplication problem, it does not solve the circulation one.
The arrangement that works is the same as for any topic group. A hub page presents the service and lists the areas covered, with a link to each location page. Each location page links back to that hub, with a descriptive anchor rather than a « back » label. And neighbouring location pages cite each other when proximity means something to the reader: « we also work in the next town » is a useful sentence, not a device.
What to avoid, on the other hand, is the automatic block listing all thirty towns at the bottom of every page. It is spotted immediately, it dilutes every link, and it turns the footer into a directory. The general linking rule applies: a few chosen links beat an exhaustive list.
Two details that make the difference in practice:
- A location page must be reachable in two clicks from the homepage. If it is only linked from the sitemap page, it is an orphan in practice.
- The link must be in the content, not only in the menu. A dropdown listing the towns is repeated on every page, so it distinguishes nothing.
What our audit measures
Several checks bear directly on this, across two categories. In on-page SEO, holding 19 checks, the audit looks at duplicate titles across all crawled pages, missing or duplicate meta descriptions, and the length of internal page titles. In targeting and keywords, holding 8, it looks for pairs of pages targeting the same query, which is exactly the symptom of duplicated location pages.
The structure and content category separately measures the median word count of your internal pages and counts those under 300 words. A run of empty location pages shows up there immediately.
What the audit does not do: it crawls up to 30 pages, not an entire website. On a site with a hundred location pages, it measures a trend on a sample, not a full inventory.