Multi-location SEO without competing with yourself: how many pages, and which to hold back

The danger in a multi-location estate is not Google. It is the page next door. Build forty near-identical town pages and the worst harm is self-inflicted: they split each other’s signals, and the ones you should never have published drag down the ones you should. The skill is deciding how many pages to build, and which to hold back.

Cannibalisation is competing with yourself

When two of your own pages chase the same intent, Google has to choose between them. It rarely chooses well. It may rank the weaker one, alternate between them, or rank neither with conviction. Picture two pages targeting “boiler repair Lewes” and “Lewes boiler repair”, near-synonyms Google reads as one intent, and it cannot tell which one should rank. You have not doubled your coverage. You have halved your authority and asked the algorithm to break the tie.

How many pages is the right number

There is no universal count, but there is a test. Build a page for a place only when you can say something true and specific about working there that you could not copy onto the next town over. A real service radius, a job you have actually done, a parking or access quirk, a local regulation. If the only thing that changes between two pages is the place name and an adjective, that is one intent wearing two costumes, and one page should carry it.

Which pages to hold back

Hold back any page you cannot defend. Thin, templated location pages are close to what Google’s spam policies describe under doorway abuse, pages targeted at specific regions or cities that funnel users to one page, and under scaled content abuse, where pages are generated mainly to rank rather than to help the person who lands there. A page for a town you have never worked in, with nothing local to say, is the clearest candidate to keep in drafts until you have something real to put on it.

• Publish a page when you can defend it with specifics a competitor could not lift wholesale.
• Hold back pages for areas you only hope to serve. Aspiration is not evidence.
• Consolidate near-synonym intents onto one strong page rather than splitting them across two weak ones.
• Fewer pages that each earn their place beat a long tail that competes with itself.

Build a hub, not a flat sprawl

The cleanest way to stop pages competing is to give each one a distinct job. The hub owns the service intent. Each spoke owns one local intent. Internal linking for local SEO covers how to wire the links so they pass the right signals, with the hub pointing down to the towns it serves and each town pointing back up to the service it belongs to.

Where Townsmith fits

Local Pages Engine is built around this restraint. It runs entirely in PHP on your own server. It makes no external calls and has no AI. Before you publish, it measures how similar each page is to its closest sibling so you can catch two pages chasing one intent, and it gives an advisory doorway-risk verdict across the whole estate so you can see which pages to hold back. The how-many-location-pages guide works through the count in detail, the duplicate content guide covers the similarity side, and the quality score documentation sets out exactly what the engine measures.