Local landing pages: how many, and when they start hurting

Every trade business gets the same pitch. Forty city pages, one for each town in your service area, and a promise of coverage across the metro.

What arrives is forty pages of identical copy with a place name find-and-replaced, each one thin, none of them ranking, and all of them now representing your business to anyone who lands on one.

How many local pages should I actually have?

As many as you have distinct things to say, which for most businesses is between three and eight.

The test is simple and it survives every algorithm change: could this page have been written by someone who has never worked in that town? If yes, it should not exist.

A page that passes the test contains things only you could write. The neighborhoods you actually work in and what makes them different, a job you did there described concretely, photos taken at that address, the local permitting or parking or access quirk your crews know about, reviews from customers in that area, and directions or timing that reflect reality.

Start with the towns that already produce revenue, not the towns you wish produced revenue. A page for a place you have never worked has nothing true to say, which is precisely why it will be thin.

If you cannot fill six of those elements for a town, do not build the page yet. Build it after you have done three jobs there and have photos.

What on a local page actually scores?

Less than the pitch implies, and the parts that score are not the parts a template farm produces.

Whitespark's 2026 Local Search Ranking Factors report had 47 local search experts score 187 factors. Three on-page items make the top twenty for local pack results, and they are all at the bottom of it: keywords in the landing page title tag at 146 (seventeenth), geographic keyword relevance of the content at 142 (eighteenth), and keywords in the landing page headings at 135 (twentieth). Compare that to primary category at 227, proximity to the searcher at 225, and being open at the time of search at 189. Darren Shaw's summary of what shifted in the 2026 edition is that review and behavioural signals rose while on-page and link signals came down slightly.

Now look at what the same panel scored near the bottom of all 187 factors, because these are the deliverables a city-page package invoices for: length of the title tag at 26, presence of an XML sitemap at 28, keywords in meta descriptions across the entire website at 33. A forty page build produces forty title tags, forty meta descriptions and a bigger sitemap. Every one of those is a countable artifact scoring in the bottom band.

Geographic keyword relevance of content at 142 is the one that pays, and it is the only one on the list you cannot manufacture by find-and-replace. It is the page that names the neighborhood, describes the job, and shows the photograph.

There is a limit on the tactic in the other direction too. The peer-reviewed generative engine optimization paper presented at KDD 2024 tested nine content tactics and found keyword stuffing was the only one that scored below the do-nothing baseline, 17.7 against 19.3. That was a simulation over informational queries with no local commercial queries in it at all, so do not overclaim. It is one more measurement where stuffing a place name into copy loses.

When do these pages start doing damage?

When they stop being distinct, they take three things from you.

They compete with each other. Ten near-identical pages about the same service leave a search engine to choose between them, and it often chooses none. Your one strong service page would have done better alone.

They make your site look automated. A visitor who lands on a page naming their town, then notices the copy is generic, has learned that you produce filler. That is a conversion problem, not just a ranking one.

They consume the attention you do not have. Forty pages need forty sets of updates when your phone number, hours, or service list change. Most businesses update the homepage and forget the rest, which is how you end up with contradictions across your own site. That is the practical half of whether NAP consistency still matters: every page is another place your details can go stale.

There is a fourth cost, and it lands on thin pages first. SparkToro's analysis with Similarweb panel data found that 68.01% of US Google searches ended without a click in 2026, up from 60.45% in 2024, and that per 1,000 US Google searches only 276 clicks now reach the open web against 374 two years earlier. Google has publicly disputed that methodology and a competing measurement using a different definition puts zero-click far lower, so use the direction and not the decimal. The direction is that fewer results pages send a click at all, and the pages that lose first are the ones that only restate the search term. That is worth thinking about alongside what happened to click through rates after AI Overviews.

The fifth cost is a policy one, and it is new. The 2026 report added a set of negative and suspension-risk factors including overlapping service areas across multi-location profiles and setting service areas larger than two hours of driving distance. City page farms and inflated service areas tend to arrive in the same proposal, from the same vendor, in the same month.

What makes a local page actually rank?

Being genuinely about that place, plus enough of the rest of your site being solid to support it.

The page needs a clear service focus, not a list of everything you do with a town name on top. One town, one service, is a stronger page than one town, twelve services.

It needs proof from that area. Photos, a described job, reviews from customers nearby. A steady review flow feeds this more than most owners realize, which is why a review system that survives staff changes is worth building before the pages, not after.

It needs internal links from somewhere real on your site, and it needs to be linked to as a normal page rather than parked in a footer list of forty towns.

And it needs the profile behind it to be right, because a page cannot outrank the map pack for a local search. The gap in the scores says it plainly: 227 for the category against 142 for geographic content relevance. If you are not appearing in the pack, look there first, and what your competitors' categories reveal is a faster read than any page project.

How do I know whether the ones I have are working?

Where does this advice break down?

In two places, and the second one is getting more expensive.

When the town is real but the page is young. A page with three months of history and no contacts is not necessarily a failed page, it is an unmeasured one. Give a genuinely distinct page two quarters before you judge it, and judge it against the towns where you actually work rather than against your best performer.

When the search is being taken by paid units before anyone reaches an organic result. Sterling Sky's State of Local SEO 2026 documents local pack ads rising from about 1% of tracked mobile reports in early 2025 to about 22% by December 2025, and Local Services Ads from about 11% to 31% of tracked queries. Joy Hawkins's read is that "Google is definitely going pay-to-play." In a market like that, a beautifully written town page is competing for a smaller organic slice than it would have two years ago. That is an argument for fewer, better pages and a serious look at the profile, not an argument for forty more.

What to do this week

Pull the list of every local page on your site with its impressions, clicks, and contacts for the last quarter. One row per page.

Sort by contacts. Mark each one: earning, fixable, or delete. Be strict. A page with impressions and no contacts for two quarters is not earning.

Take your top two towns by actual revenue and add what is missing to those pages. Photos from real jobs, a described project, nearby reviews, the neighborhoods you serve there.

Redirect the delete list to the relevant service page. Do it in one batch and note the date.

Check your service area settings on the profile while you are there, and pull back anything beyond two hours of driving.

Questions about how many pages to build?

Email me at eric@seod.com with the towns you serve and, next to each, roughly how many jobs you did there last year. I will send back which towns justify a page now, which ones need three more jobs before they will, and which ones you should stop planning for entirely.

I do this myself and it takes about fifteen minutes. If you already have pages built, send the list of URLs instead and I will tell you which ones I would keep, rewrite, or redirect.

The test is whether you have something true to say about the town. Search visibility builds service and city pages on that basis, to answer a real search rather than to pad a word count. There is more on local search in the library.