A plumbing company sells drain cleaning. Should that be one page, or should there be separate pages for drain cleaning, drain unblocking, blocked drains, and emergency drain service? The keyword tool lists all four with respectable volume, and the temptation is to build all four.
The number of pages a service needs is not decided by how many ways people phrase it. It is decided by how many distinct results pages Google returns for those phrasings.
What actually decides the number of pages?
One landing page per distinct search intent, not per keyword and not per service name. If two queries produce substantially the same top ten results, Google has already told you it considers them the same question, and two pages will compete for one slot.
This is the test, and it takes a couple of minutes per pair. Search both queries, list the top ten URLs for each, and count the overlap. Heavy overlap means one page. Different result sets — different page types, different businesses, different content shapes — mean two.
Which signals justify a second page?
The ones that describe a different buyer or a different job, rather than a different vocabulary for the same one.
| Signal | Build a second page? | Why |
|---|---|---|
| Different results in the top ten | Yes | Google is treating them as separate questions |
| Different urgency (emergency vs planned) | Usually | Different buyer, different proof, different call to action |
| Different price band or process | Usually | The page has to answer different objections |
| Synonym or plural of the same term | No | Same intent; one page can rank for both |
| Same service, different city | Only with real local substance | Otherwise it is one page duplicated |
| Volume in a keyword tool | No | Volume is not evidence of a separate intent |
The emergency row is the one worth dwelling on. "Blocked drain" and "emergency drain unblocking at 2am" look like the same service, and commercially they are two different products: one buyer is comparing quotes, the other is not comparing anything. That is a real second page.
A page earns its existence by answering a question no other page on your site answers. If you cannot state that question in one sentence without using the words on another page, you are building a duplicate with a different title tag.
What goes wrong when you split too far?
The pages get thin, and the internal links that would have made one page strong get spread across four that are not.
The mechanism is worth naming because it is invisible on a page-by-page review. Page authority flows through internal links, and a site has a finite number of them pointing at any topic. One page receiving forty contextual links competes differently from four pages receiving ten each. Splitting also multiplies the maintenance: four pages of near-identical copy drift apart, and by the second year one of them is quoting a price that changed.
Most split-too-far pages end up as orphan pages anyway. Nobody links to variant three from the navigation, because the navigation would look ridiculous with four drain-cleaning entries. A page nothing links to is a page the site itself does not believe in.
What goes wrong when you do not split enough?
One page tries to satisfy two different readers and gives each of them half a page.
A combined page for residential and commercial work has to open with something. Whatever it opens with, the other audience is now reading copy addressed to someone else. The on-page SEO problem follows from that: the title and headings can only make one promise, so the page matches one query well and the other approximately. Approximate matches lose to specific ones.
Building the page structure from the internal org chart. Service lines that are separate departments internally are frequently one decision to the buyer, and departments the business treats as one are frequently two. The buyer's view is the one search reflects.
Where do location pages fit?
They multiply everything above, which is why they need a higher bar. Four services across twelve towns is forty-eight pages, and forty-eight pages of interchangeable copy is a maintenance liability that ranks for nothing.
The bar that works: a location page is justified when it carries something that could only be true of that location. Named projects, local pricing, a real address, staff who work there, permit or regulatory detail specific to the area. Where you have that, the page has a reason to exist. Where you have only the town name swapped into a template, one strong regional page will do more.
How do you audit the pages you already have?
Group them by the query they were built for, then check what search actually did with them. Three passes, in this order:
- Pull impressions per page. Pages with near-zero impressions on their target query were never in the auction and are candidates to merge.
- Look for pages trading positions on the same query. If two of your URLs alternate in the results for one term, they are splitting the signal and one should redirect to the other.
- Check internal links. Any page with no links pointing at it has not been tested yet — link to it properly before deciding it failed.
When you merge, move the genuinely distinct sections onto the surviving page, redirect the old URL, and repoint the internal links so they land on the target directly rather than through a hop.
Take the service you have most pages for and run the overlap test on its two closest queries. If the results are the same set, you have your first merge, and it will make the surviving page stronger than either was.
Not sure whether to split a page or merge two?
We will map your services against the queries they actually compete for and tell you which pages to build, merge, or leave alone.
Get in Touch →Does this apply to your site?
Reading about it is one thing. Point the scan at your own site and see whether this applies to you, and what it is worth fixing.
Free and unlimited. No account, no card, and you get every finding rather than a teaser.