A page called "Plumber in Kettering" is either the most useful page on a plumber's website or the reason Google stops trusting the site, and the difference is whether anything on it is true of Kettering.
Most of the town pages we audit for trade businesses around Northampton are the second kind.
The idea behind them is sound. A heating engineer in Northampton also works in Wellingborough, Daventry and Towcester, and people in those towns search "boiler repair Wellingborough", not "boiler repair Northampton". The home page cannot answer every town, so a page per town sounds right.
The problem is the shortcut. One page gets written, the town name gets swapped eight times, and the site now has nine pages that say the same thing. Google has a name for that, it has published the policy, and it has been catching the pattern since 2015.
What Google's doorway policy says
Google's spam policies define doorway abuse as pages built to rank for near-identical searches without giving the visitor anything different. The exact wording, from the spam policies page last updated in August 2026, is short enough to quote in full:
"Doorway abuse is when sites or pages are created to rank for specific, similar search queries." Google Search Central, spam policies
The examples under it are the part that matters for a trade business. One is "multiple domain names or pages targeted at specific regions or cities that funnel users to one page". Another is pages that exist only to push visitors into the useful part of the site. A third is near-duplicate pages that sit closer to the search result than a browsable site structure does.
Read those against a typical set of town pages. Eight pages, each one an intro paragraph with the town name, the same list of services, the same call-to-action, no photos, no jobs, no reviews from that town. Every example in the policy applies. The page is a doorway to the contact form, and nobody would stop on it.
It is not only trades that fall into this. A law firm that publishes a page for every town it hopes to take instructions from, with no office and no real case in any of them, has built the same doorway under a smarter logo.
Google's original announcement of the policy in March 2015 came with a list of questions site owners should ask themselves. Impression Digital reproduced them at the time, and they are still the clearest test I know of:
- Is the purpose of the content to please search engines and funnel traffic into your site, or is it part of your site's experience for users?
- Does the content deal with specific subject matter, yet get optimised for generic terms?
- Do the pages reproduce information found elsewhere on the site just to draw more organic traffic?
- Do these pages exist as an island, hard to find from the rest of the site?
- Are the links to them created just for search engines?
A page that fails questions one, three and four is a doorway page whatever its title says.
The test: would this page exist if Google did not?
The quickest way to judge a service-area page is to ask whether the business would have built it with search engines switched off. Google's own helpful content guidance asks the same question less bluntly: "Is the content primarily made to attract visits from search engines?"
Imagine the plumber hands a customer in Wellingborough a printed leaflet. Would that leaflet say "we are proud to serve Wellingborough and the surrounding areas" followed by the same list of services as every other leaflet? No. It would say "we did the boiler swap on the new estate off Doddington Road last month, here is a photo, here is what it cost, we are 20 minutes from you on the A45".
That second leaflet is a service-area page. The first is a doorway. The content that makes the page useful to a human is exactly the content that makes it distinct to Google, so there is no conflict between the two. There is only the work of gathering it.
Google's guidance has a second question worth keeping to hand: "Does your content leave readers feeling like they need to search again to get better information from other sources?" If a Wellingborough resident reads your Wellingborough page and still does not know whether you cover their postcode or what you have done nearby, they will go back to Google. That is the signal Google reads.
What makes a town page real
A service-area page earns its place when it carries proof that the business works in that town, and nothing else does the job. There are five kinds of proof, and each one is something only this business can supply.
Jobs done there. Two or three recent jobs, each with one line on the problem, what was fitted, and roughly when. Three sentences each is enough. The Search Engine Land guide by Miriam Ellis puts project write-ups top of the list of page-level content.
Photos from those jobs. A photo of a van outside a semi in Rushden is worth more to a reader than a stock image of a smiling engineer, and it is impossible to reuse on the Corby page.
Reviews that name the town. A Google review that says "came out to us in Daventry the same afternoon" belongs on the Daventry page. Reviews that do not name a place belong on the home page.
Travel time. "About 25 minutes from our base in Northampton" is a fact a customer wants and a copy-and-paste template cannot produce. Google's Business Profile guidelines say a service area "shouldn't extend farther than about 2 hours of driving time" from where the business is based, which is also a sensible outer limit for which towns get a page at all.
Local detail that is true. If you genuinely find the same 1970s back boilers across one estate, say so. If the only local detail you can think of is the name of the market square, leave it out. Landmark-dropping is the most recognisable doorway habit after the town-name swap, and readers in that town can tell.
BrightLocal's guide to service area pages gives a worked example of the failure mode: fifteen plumbing pages that are identical apart from the city, and it advises mentioning the town only a handful of times on the page. If a page cannot be written without repeating the town name in every paragraph, there is nothing else on it.
Before and after
The difference between a doorway and a service-area page is visible in the first screen. Below is a Kettering page as we typically find it, against the same page rebuilt.
| Element | Doorway version | Service-area version |
|---|---|---|
| Opening | "Looking for a reliable plumber in Kettering? Our team of experienced plumbers serves Kettering and surrounding areas." | "We cover Kettering from Northampton, about 25 minutes on the A43. Most Kettering call-outs are in Barton Seagrave and Burton Latimer." |
| Services | The same bulleted list as every other town page | The three services actually booked most often from Kettering, with prices |
| Proof | None | Two jobs from the last three months, each with a photo and one line |
| Reviews | Site-wide star rating | Two Google reviews that mention Kettering, quoted with dates |
| Local detail | "Kettering is a historic market town in Northamptonshire" | Nothing, unless there is a real pattern worth telling customers about |
| Links to it | Footer only, with seven other towns | Linked from the areas page, the home page, and the two job write-ups |
| Would it exist without Google? | No | Yes, it is the page you would send a Kettering enquiry |
The right-hand column is harder to produce. It needs a photo folder, a review export and ten minutes with whoever did the jobs. It also cannot be made for a town where the business has never worked, which is the point.
Service-area pages and the map pack
A town page does not move your map pack position in that town, and it is worth being clear about that before anyone builds nine of them. Google's local ranking help page says local results rest on relevance, distance and prominence, and distance means "how far each business is from the customer who's searching".
For a service-area business with its address hidden, Google still uses the verified address for distance. A long-running Local Search Forum thread on exactly this question concludes, in the words of the accepted answer, that "service areas are not a ranking factor at all". That is forum opinion, and Google has not documented it either way, but it matches what we see: a Northampton plumber does not appear in the Kettering map pack because of a Kettering page.
Where the page does its work is the organic results underneath the map, and in AI answers. Whito's August 2026 study of 159 Google AI Mode answers to "recommend a trustworthy [trade] in [town]" across 20 UK towns found the business's own website among the sources for 90% of answers. Checkatrade appeared in 83% of the answers for electricians, plumbers and roofers. The town pages that get cited are the ones with something specific to quote, because in AI answers third-party mentions often matter more than your own site.
How many pages a Northampton trade business should have
These town pages sit alongside the service pages in a small business SEO structure. The rule we use is one page per town where the business can show at least three jobs completed in the last twelve months, with photos, and at least one review that names the town. If a town does not meet that bar, it goes on a single "areas we cover" page with a map and a travel-time list, and nothing more.
For a one-van firm based in Northampton, that usually means the home page covers Northampton itself and three to five town pages cover the places that actually generate work: typically some mix of Wellingborough, Kettering, Daventry, Towcester, Rushden and Brackley. Corby and Milton Keynes get pages only when the job log says they should. Villages never get their own page. They get a line on the nearest town's page.
Two further limits. The Search Engine Land guide suggests that with ten or fewer town pages they can sit in the main navigation, and beyond that they need a sitemap page, which for a small trade business is a sign there are too many. And every page needs re-checking once a year, because a 2024 job photo on a page dated 2026 reads as abandoned. If you cannot keep six pages current, build four.
The same rule tells you when to delete. A town page with no jobs added in a year, no review naming the town and no enquiries in Search Console is a doorway that happens to be old. Fold it into the areas page and redirect it.
Where this leaves the eight-page site
An existing set of town-swapped pages has three honest fixes. Rebuild each page with real proof, merge the lot into one areas page and redirect the old URLs, or keep the two or three towns that have evidence and merge the rest. Leaving them up and adding two more, which is what most sites do, is the one option that gets worse every year.
The question to put to every town page is whether you would send it to a customer who lives there. If the answer is no, the page is working for Google, and Google has said it does not want it. Where this fits in a full retainer is covered in our guide to what a UK SEO agency should cost and deliver. That question is the starting point for our local SEO work with trade clients. If you want us to look at your town pages and say which ones should stay, get in touch.