Location pages built with AI

The test every location page has to pass

Remove the town name from the page. If what remains could belong to any other town you serve, the page has no reason to exist and adding more words will not give it one.

That is the same measure we apply to our own pages, with the distinguishing token stripped before anything is compared. It is a harsher test than it sounds and it is the honest one.

Our page on doorway pages covers what happens to sets that fail it. The short version is that they compete with each other, and a search engine or an answer engine treats the set as a single thin page.

What is genuinely different about a place

The work itself. Older housing stock, particular soil, a common local fault, a regulation that applies in that county and not the next. Every trade has these and they are invisible from outside.

The logistics. How long you take to get there, which days you are in the area, whether there is a call-out charge, where you park, what the access is usually like.

And the evidence. Jobs you have actually done there, with photographs and, where you have permission, the street or neighborhood named. Nothing substitutes for this.

How many pages to build

As many as you have something to say about. That is usually far fewer than the number of towns in your service area, and building only those is a stronger site than building all of them.

One page covering the wider area is better than twenty empty town pages. A visitor finds what they need either way, and the single page can be genuinely good.

Add towns as you accumulate work in them. Our page on location pages covers the shape, and a page added when you have three real jobs to describe will always beat one added speculatively.

What every location page needs

The list below is the minimum for a page that earns its place. If you cannot supply at least four of these items for a town, that town does not need a page yet.

• Work you have actually done there, described specifically
• Photographs from jobs in that area
• Something true about conditions or buildings there
• How you cover it, including timing and any charge
• The neighborhoods or landmarks people use locally
• A named contact route for that area
• Anything local that changes how the job is done

Generating them safely

Supply the differences yourself, one town at a time, and let generation handle the structure. That ordering works. Asking for twenty location pages from a list of names does not.

Check the coverage list against reality before publishing. Generated service areas routinely include towns that sound plausible and that you do not serve, which produces wasted calls.

And check the pages against each other, not only individually. Our site check reports duplication across a site, which is the failure that only shows up in comparison.

Linking a location set together

Every location page should link up to a parent page listing the whole service area, and that parent should link down to each one. A flat pile of location pages with no parent is hard to crawl and harder to navigate.

Link sideways sparingly. Neighboring towns are worth connecting where a reader might genuinely be between them, and a full grid of links between every pair reads as engineered.

Our page on site structure covers the shape, which matters more for location sets than for almost any other kind of page.

Keeping them true

Coverage changes. A town you stopped serving still has a page saying you do, and that page will keep producing calls you have to turn down until somebody notices.

Review the set whenever your area changes, and delete rather than leave stale pages live. A removed page with a sensible redirect is better than a page that lies politely.

Add new work as it happens. A location page that gains a real job description every few months stays genuinely different, which is the only durable way to keep a set like this defensible.