Location pages that earn their rankings

Every multi-location business eventually meets the same tempting shortcut. You rank in Sydney; you'd like to rank in Wollongong; therefore you generate forty pages, one per suburb, each saying "We offer excellent [service] in [suburb]" with the suburb name find-and-replaced. It takes an afternoon. It feels like strategy. And it is exactly what Google's doorway-page guidance exists to demote — pages that exist for search engines rather than humans, funnelled into one real destination.

And yet location pages, done honestly, are some of the best-performing pages we ship. The hospitality local-SEO work we do for venues depends on them. The difference between an earning location page and a doorway page isn't the template — it's whether the page knows something the homepage doesn't. Here's the standard we build to, and the uncomfortable question that decides which locations get pages at all.

The doorway test, stated plainly

Google's bar is roughly: would this page exist if search engines didn't? I'd sharpen it. For every proposed location page, ask: what does someone in this place need to decide that our homepage can't tell them? If the answer is a list — we'll get to what goes on it — build the page. If the answer is "nothing, we just want to rank there", you've described a doorway page and saved yourself the hosting.

This test kills most mass page programmes before they start, which is the point. Twenty strong location pages beat two hundred thin ones in rankings, in conversion, and in the survival rate through core updates. Fewer, truer pages is not a compromise. It's the strategy.

The unique-value checklist: what a real location page knows

A location page earns its URL when it carries information that is genuinely local, genuinely specific, and genuinely decision-relevant. Our checklist — a page needs at least four of these seven to justify existing:

Real availability. Which services, menus or products this location actually offers — not the company-wide list with the suburb name stapled on. If the Wollongong studio doesn't do installations, the page must say so.
Local proof. Reviews, projects or customers from this area, with specifics. "We've fitted out fourteen cafés on the north shore, including three on this street" is locational evidence no template can fake. No local proof yet? That's a strong signal this location isn't ready for a page.
The humans. Names, photos and single-line bios of the people a visitor would actually meet. For service-area businesses, the technician who covers the region. This single element does more trust work than any other, and it's the one doorway pages never have.
Practical logistics. Parking, access, public transport, service radius with real boundaries, actual opening hours including the local public-holiday quirks. Decision-relevant detail is what separates a page for humans from a page for crawlers.
Local inventory or pricing friction. If prices vary by market, say so and show them. Hiding it doesn't make it universal.
Genuinely local editorial. The neighbourhood content that's real: "we service the heritage terraces in Paddington — here's what council approvals mean for your timeline" is expertise placed somewhere. It can't be generated once and swapped forty times, which is precisely why it works.
Local structured data. Location-specific LocalBusiness markup with real NAP details, hours and geo — the schema playbook covers implementation, but the rule here is: mark up only what's on the page.

Below four, the honest move is usually one strong regional page ("/service-areas/illawarra/") covering several suburbs properly, rather than four hollow ones.

Templating with real data

"Don't use a template" is bad advice — templates are how you keep forty real pages consistent and maintainable. The rule is: template the structure, never the substance. The template owns the layout, the components, the schema scaffolding and the internal-link modules. Every content slot must be filled from per-location data, and the build should refuse to render a slot that's been left at its placeholder.

In practice: store location data as structured records (hours, staff, services, proof points, photos, reviews), and make the page a view over that record. When a record is thin, the page doesn't ship — the location joins the regional page instead. This inverts the usual failure mode where publishing is the default and quality is policed afterwards. We apply the same discipline to programmatic SEO generally: the pipeline must have a quality floor it cannot fall below, or it will.

Two more template rules that save you later:

One canonical pattern, no duplicates. Never publish both /sydney/ and /locations/sydney/; never let city and suburb pages target the same query. Map one page per distinct search intent before you build anything.
The homepage doesn't dodge the work. Location pages convert when the homepage links to them as destinations ("Find your local studio") rather than as an SEO afterthought in a footer list. Footer-farm link blocks are a doorway-page tell.

Internal linking: the structure that signals substance

Location pages live or die on how they join the site. Our standard pattern:

A locations hub (/locations/) that's a real page — map, regions, genuine guidance on choosing — not a naked grid of links.
Region intermediaries where volume justifies them: city → suburb. Never deeper than three levels.
Cross-links to relevant case studies and services. A location page linking to a project completed in that area is the strongest topical signal you own; this is internal linking as architecture, not as decoration.
Breadcrumbs with markup, so the hierarchy is legible to humans and crawlers alike.

What to avoid: reciprocal link exchanges with every location page linking to every other, and the classic "popular searches in other areas" module that links forty suburbs from every page. Both patterns scream template. If every page links everywhere, nothing is a structure.

The helpful-content lens

Since the helpful-content system merged into core ranking, site-level signals matter: a section of doorway pages can drag down pages that had nothing to do with them. This changed the risk calculus. A programmatic location programme is no longer "maybe it works, maybe Google ignores it" — it's a liability you hold on the balance sheet.

So run the annual review: for each location page, check impressions, clicks, and — more honestly — whether the page has earned a single enquiry. Pages that haven't earned one in a year get improved with real local substance, merged into their region page, or removed with a redirect. Treating content pruning as a lever applies with double force to location pages, because their decay is invisible until a core update makes it very visible indeed.

When fewer pages win

Three scenarios where we actively recommend against location pages: service businesses whose work is genuinely identical everywhere (one strong service page plus a service-area page beats twenty clones); young businesses without local proof yet (earn three projects in an area, then the page writes itself); and markets where the search volume doesn't exist (ranking #1 for a query nobody makes is a participation trophy). In all three, the growth plan is better served by putting that effort into the reviews, photography and case studies that will make future location pages undeniable.

The shortest version of all of this: build the page for the customer standing in that suburb with their phone in their hand. If the page helps them decide, it will rank. If it exists to rank, it eventually won't.

Key takeaways

Test every proposed location page against one question: what does a local visitor need to decide that the homepage can't tell them?
Require at least four of seven unique-value elements — availability, local proof, local humans, logistics, real pricing, local editorial, local schema.
Template the structure, never the substance; build a quality floor the pipeline can't publish below.
Link deliberately: real hub, shallow hierarchy, cross-links to locally relevant work, no footer farms.
Review annually and prune without mercy — doorway pages are now a site-level liability, not a neutral experiment.

FAQ

We're a service-area business with no storefronts — can we still have location pages? Yes, and arguably you need them more. But the bar is the same: real service boundaries, the actual team member covering the area, proof from that area, and honest logistics ("we charge a call-out fee beyond 30km" builds trust rather than losing it). "We serve all areas" pages with a swapped suburb name are doorway pages whether or not you have a shopfront.

How many location pages are too many? The number where you run out of unique substance. For most businesses that's somewhere between five and forty. Google doesn't penalise a count; it demotes patterns. If page 31 required you to invent filler, it should have been a region page.

Should each location page target its exact-match suburb keyword in the title? Use natural local phrasing — "Plumbers in Marrickville" is fine if it's true and the page is substantive. What you can't do is let the keyword be the page. Titles follow substance; they don't substitute for it.

What do we do with legacy doorway pages that used to rank? Improve, merge or redirect — in that order of preference. Pages with remaining equity and some substance get rebuilt with real local data; clusters of thin pages get merged into stronger regional pages with 301s; the irredeemable get pruned. This pairs naturally with a technical SEO pass so redirects and canonicals stay coherent.