Every cheap SEO package sells the same thing to local service businesses. One page per city, one page per service, and then one page per service per city. Garage door repair in Chula Vista. Garage door repair in La Mesa. Garage door installation in Chula Vista. Forty pages, sometimes four hundred, mostly identical except for a swapped city name.
The logic sounds airtight: more pages, more keywords, more chances to rank.
The result is usually that you rank for less than you did with five pages. Not because Google punished you, but because you spent six months making it hard for Google to answer a simple question.
01Google only picks one
For any given search, Google shows one page from your site. Not your best page. One page.
When five of your pages are built around the same intent, Google has to choose between them, and it has almost nothing to choose on — the pages are near-identical by design. So it picks something. Often the page that happens to have an internal link pointing at it, or the one that got crawled most recently. It's not a considered decision, because you didn't give it anything to consider.
That's keyword cannibalization. The pages aren't competing with your competitors. They're competing with each other.
"Five pages targeting one search don't get five chances to rank. They split the authority that one strong page would have had, then hand Google the job of picking a winner at random."
The compounding part is worse than the split. Every link you earn, every internal link you place, every bit of accumulated relevance gets divided across the group instead of concentrated in one place. A single page with all of it might have cleared the top three. Five pages holding a fifth each clear nothing.
02The symptoms
Cannibalization looks like a plateau, which is why it gets misdiagnosed as "SEO takes time."
- Impressions climb, clicks don't. More pages get seen. None get clicked, because none rank high enough to be clicked.
- The wrong URL shows up. You search your main service and Google returns a thin city page instead of the service page you actually invested in.
- Rankings swap week to week. Position 8 on Monday from one URL, position 14 on Friday from another. That flickering is Google re-picking, and it's the clearest tell there is.
- A page you deleted still outranks the one that replaced it.
- Your best page slipped right after you published a batch of new ones. People almost always read this as a coincidence.
Open Performance, go to the Queries tab, and click a query you care about. Then switch to the Pages tab with that query still filtered. You're now looking at every URL on your site that has ranked for that one search. One URL and you're clean. Four and you've found it. Do this for your five highest-value searches and you'll usually have the whole picture in twenty minutes.
03When location pages actually work
This isn't an argument for one page and nothing else. Location pages work — they just have to earn their existence.
A page for Chula Vista is worth building when there's genuinely different content on it. Jobs you've done there, photos from those jobs, a real service area boundary, response times, a permit quirk or an HOA rule that applies there and not elsewhere, a review from a customer on that street. That page can rank, because it answers something the others can't.
A page that is the Chula Vista version of the La Mesa version of the El Cajon version is not a page. It's the same page with a find-and-replace, and Google has been able to spot that for well over a decade.
The honest test: could a customer tell which city page they landed on with the city name covered up? If not, it shouldn't be its own URL.
04Untangling it without losing ground
The instinct is to delete everything. Don't — you'll throw away links and rankings you can keep.
- Group by intent, not by keyword. Everything that answers the same customer question goes in one pile, however differently it's worded.
- Pick the survivor in each pile. Usually the page with the most links or the strongest existing position, not the newest or the prettiest.
- Merge the useful parts in. Anything genuinely good on the losing pages moves onto the survivor before anything gets removed.
- 301 the rest to it. Redirect, don't delete. The accumulated value transfers; deleted pages just drop it.
- Fix the internal links. This is the step everyone skips and it's most of the effect. Every internal link pointing at a service should point at the same URL.
- Then wait properly. Re-crawling and re-consolidating takes weeks, not days. Judging it after ten days tells you nothing.
Where this lands
The businesses that win local search rarely have the most pages. They have fewer pages that each answer one thing completely, and enough internal linking discipline that Google never has to guess which page is which.
Adding pages is easy to sell because it's easy to count. Forty pages is a deliverable. One page that finally ranks is harder to put on an invoice and worth considerably more.
If your traffic went sideways after a content push, this is the first thing to rule out — and it's usually the thing. We ran this exact exercise on our own site and cut eight city pages down to one.
Want to know if this is happening to you?
Our Measure plan includes Search Console monitoring and monthly SEO work driven by what it actually shows. $199/month.