Are Service Area Pages Doorway Pages? Google's Actual Rule

Service area pages are not doorway pages by default. Google's spam policy defines doorway abuse as pages "created to rank for specific, similar search queries" that "lead users to intermediate pages that are not as useful as the final destination," and it names one example directly: "having multiple domain names or pages targeted at specific regions or cities that funnel users to one page." Notice what triggers that description. It is not the existence of a page per city. It is that the pages have nothing on them, so the visitor's real destination is somewhere else. A city page that tells a homeowner in that city something true and specific about how you work there is a useful page for a local query, and Google treats it as one.

The practical problem is that most city page projects fail this test by accident. Somebody writes one solid page, duplicates it eighteen times, swaps the town name in the heading and the first paragraph, and publishes the set on a Tuesday. That is the exact pattern the policy describes, and it is why so many service businesses conclude that city pages "do not work anymore." They work. Duplicates do not.

What Google actually says, in its own words

Two policies matter here, and they are worth reading rather than paraphrasing, because both are narrower than the folklore around them.

Doorways. Google's spam policies describe doorway abuse as sites or pages "created to rank for specific, similar search queries" that "lead users to intermediate pages that are not as useful as the final destination," and list "having multiple domain names or pages targeted at specific regions or cities that funnel users to one page" as an example. The operative word is "funnel." The policy is aimed at pages whose only job is to catch a query and pass the person along.

Scaled content abuse. The same page defines this as "many pages are generated for the primary purpose of manipulating search rankings and not helping users," including "using generative AI tools or other similar tools to generate many pages without adding value for users." Again, read the qualifier. Google has stated plainly that appropriate use of AI is not against its guidelines and that content is judged on quality, originality, and usefulness rather than on how it was produced. The tell is the absence of value, not the presence of automation.

Put those together and you get a test you can apply to any page you are about to publish: would a person who lives in this city learn something here that they could not have learned on the homepage? If yes, you have a service area page. If no, you have a doorway, regardless of how it was written or how many words it has.

Doorway page or service area page: the honest comparison

How many service area pages should you have?

Most US service businesses land somewhere between 15 and 30, and the number should come from your books rather than from a map. The cities worth a page are the ones that already produce paying jobs, which you can pull from a year of invoices, your CRM, or your ad reporting. If you are running several channels and cannot see which places convert, a dashboard that unifies your ad channels and revenue in one place answers the question faster than exporting three CSVs and reconciling them by hand.

There is also a useful outer boundary that comes from Google itself. A service area business can list up to 20 service areas on its Google Business Profile, those areas must be named as cities, postal codes, or defined areas rather than a radius, and Google says the boundary of your overall area should not extend more than about two hours of driving time from where the business is based. That is guidance for your profile, not your website, but it is a sane sanity check for your page list too. If you would not actually drive there, you probably cannot write anything true about it.

A rough shape by business type:

• One van, one metro: 5 to 10 cities, the suburbs that already call you.
• Small crew, full metro: 15 to 25 cities, taken from last year's paid jobs sorted by city.
• Multiple crews, several counties: 20 to 40, grouping small towns into regional pages instead of splitting hairs.
• Storefront plus a service radius: the location page, kept primary, plus 8 to 15 service area pages.
• Statewide or national: regions, not cities. City pages at that scale are the textbook doorway pattern.

What makes a city page genuinely different

This is the part people skip, so here is the concrete list. You do not need all of it on every page. You need enough that the page could not be swapped with the one next to it.

Coverage detail. The neighborhoods, subdivisions, or postal codes you actually work in, and honestly which parts of the city you do not cover. Admitting a limit is one of the strongest trust signals a local page can carry.

Response reality. How long it takes you to get there, whether you have same-day availability in that area, and which days your crew is normally nearby. A homeowner with a leak is reading for exactly this.

Local conditions. Housing stock and its age, common construction types, the seasonal problem that shows up there, water hardness, soil, tree cover, whatever is real for your trade. A roofer in a hail corridor writes a different page from one two counties east.

Permits and rules. Which municipal office issues the permit, typical turnaround, any local code quirk you handle routinely. This is the single hardest thing for a competitor to copy and the clearest sign to a reader that you have worked there.

Work you have done nearby. Named jobs, with permission, at street or neighborhood level. Photos from that area rather than stock. Reviews from customers in that city if you have them.

Service emphasis. Not the same six services in the same order on all pages. If one suburb is 70% water heater replacements and another is mostly repipes, the pages should say so.

Does the page need 500 words? 1,000?

Google has never published a word count for anything, and any number you see quoted is somebody's rule of thumb rather than a threshold. That said, the numbers are downstream of a real constraint: it is difficult to include coverage, response times, local conditions, service emphasis, and a few genuine questions in under a few hundred words. If your page is 120 words, the issue is not the count, it is that you have not said anything. If your page is 1,400 words and 1,100 of them are identical across the set, the count has not helped you.

A more useful discipline than counting: open two of your city pages side by side and read them. If you cannot tell which is which without looking at the town name, neither can Google.

Do service area pages work without an address in that city?

Yes, and this is the normal case. A service area business is defined by not serving customers at its own location, which is why Google Business Profile lets you hide your address entirely and list service areas instead. Your city pages compete in ordinary organic results, where a physical address in the town is not a requirement.

What you must not do is invent one. Renting a mailbox, using a virtual office, or listing a friend's address in a target city is a listing suspension risk and it has nothing to do with whether your web page ranks. The map pack and the organic results are different systems, and faking your way into the first will not help the second.

Can you use AI to write service area pages?

Yes, and Google has said so about as clearly as it says anything. Its guidance is that appropriate use of AI or automation is not against the guidelines, and that content is assessed on quality, originality, and usefulness regardless of how it was produced. The line in the spam policy is scaled content abuse: generating many pages without adding value for users.

So the question is not whether AI wrote the page. It is what the AI was given. Asked to spin one page into twenty variations, any tool produces doorways, and it produces them faster than a human could. Given real input per city, coverage, response times, local conditions, permits, jobs you have finished there, the same tool produces twenty pages that pass the test comfortably. The input is the whole difference, and it takes about two minutes per city to supply.

This is the workflow behind our service area page builder: the agent asks what is different about each place you serve, writes a distinct page from your answers, and links each one into your multi-page site structure and sitemap so the pages are crawlable rather than stranded. What it will not do is pretend to know what makes work in your towns different, because no tool has that.

How fast should you publish them?

In batches, over several weeks. Publishing thirty new pages in one day on a site that had twelve is a shape worth avoiding, and the practical argument is stronger than the risk argument: staging the rollout means you find out which pages attract calls before you write the next ten. Start with your five best cities, watch them for a month, then expand. If the first five do nothing after a fair trial, twenty-five more copies of the same idea will not fix it.

Common questions

Are service area pages considered doorway pages?

Not inherently. Google's doorway policy targets pages aimed at specific regions or cities that funnel users to one page without being useful themselves. A city page carrying real coverage detail, response times, local conditions, and questions specific to that area is a useful result for a local query. Twenty copies of one page with the name swapped is precisely what the policy describes.

How many city pages is too many?

The count is not the problem, the emptiness is. That said, if you cannot say something specific and true about a town, that town does not get a page. Most US service businesses find that ceiling somewhere between 15 and 30 cities. Beyond your genuine service radius, roughly two hours of driving by Google's own service area guidance, you are writing pages about places you do not work.

Should I make one page per city or one page listing all my cities?

Per city, if each one earns it. A single page listing every town you serve tends to rank for none of them, because it matches no specific local query well. For towns you cannot write meaningfully about, group them into a regional page with their neighbors rather than publishing a placeholder.

Will duplicate city pages get my whole site penalized?

The usual outcome is quieter than a penalty and just as unhelpful: the pages are filtered out of results and never rank, so the project appears to have simply failed. Google's spam policies do allow for ranking action against sites in breach. Either way, the fix is the same, which is to make the pages worth reading or to delete the ones you cannot.

What about neighborhood pages, are those different?

Same rules, higher difficulty. Neighborhood pages only make sense when neighborhoods are genuinely how customers in your market think and search, common in dense urban areas, rarely in suburban ones. The unique content bar is harder to clear at that granularity, so most businesses are better off with strong city pages first.

Do I need separate service area pages for each service in each city?

Only at real scale, and only when the demand is there. Multiplying six services by twenty cities gives you 120 pages, which is where a page set stops being a website and starts being a matrix. Build city pages first, weight each one to the services that area actually calls about, and add a service plus city page only where search demand clearly justifies it.

How long before service area pages rank?

Expect a few weeks for indexing and a few months for position to settle, longer in competitive metros. Local rankings also depend heavily on things that are not on the page: your Google Business Profile, reviews, and citations. A city page will not out-rank an established local competitor on its own, but it is the piece most service businesses are missing entirely.

Do city pages help with AI assistants and AI Overviews?

They can, for the same reason they help in search: assistants answer local questions by pulling from pages that make specific, checkable claims. A page that states which neighborhoods you cover, how fast you get there, and what you charge for the common jobs is far easier for a model to cite than one that says you proudly serve the greater metro area. Generic pages are unquotable, which is a cost that did not exist five years ago.