Classic Search
Keyword Cannibalization
Keyword cannibalization happens when two or more pages of the same website compete for the same query and intent. Search engines may not consistently favour one page, so impressions, links and clicks are split between them, and the pages often swap positions or rank lower than one stronger page would.
- Industry term
By QueryRev ResearchPublished Updated5 min read
Why it matters.
Cannibalization hides in growth. A shop publishes “platform beds explained” one year and “platform vs slatted beds” the next, and both now answer the same question. Neither gathers all the internal links or all the clicks, and the URL shown in the results changes from week to week, which makes reporting unreliable as well.
Not every overlap is a problem. Two pages can rank for one query when they serve a different search intent, such as a category page and a buying guide. It is a problem when the pages are interchangeable and their positions are unstable.
When two pages are interchangeable, consolidation is the usual fix. Google’s guidance on duplicate URLs describes consolidation as a way to combine the signals of several URLs, such as links, into one preferred URL, and names redirects as the strongest of the methods it lists[1]. For a URL that changes for good, Google recommends a permanent server-side redirect[2].
A merge helps in AI answers only if the page you keep can be read. Of the top sites in QueryRev’s study, 896 of the 6,294 whose robots.txt could be read, 14.2%, block at least one AI crawler by name[3]; consolidating pages does nothing for an answer engine whose crawler the site keeps out.
It is most common on sites that publish often and rarely prune: blogs with yearly versions of one article, shops with a category page and a filtered page for the same thing, and help centres where one question is answered in several places.
How it is measured.
In Search Console, filter by the query and list the pages that earned impressions for it. The warning signs: two of your URLs with substantial impressions for one query, positions that trade places over time, and a ranking URL that changes between days. Then read both pages: do they answer the same question for the same reader?
The Performance report shows the last three months by default[4], long enough for a swap to show as a pattern rather than one day’s noise.
Look at the clicks as well as the impressions. Two URLs that share impressions but where one takes almost every click are less of a problem than two that split both; the second case is where a merge has the most to gain.
Measured by
- URLs per query
- Position swaps
- Ranking URL changes
A worked example.
Sample data: an invented site, to show the method.
For “platform vs slatted bed”, two blog posts share 6,100 impressions a month and move between positions 6 and 11, trading places. One has the stronger history. Merging the weaker into it, with a 301 redirect and its unique sections kept, leaves one URL to collect the signals. How many of the combined impressions it keeps is measured afterwards, not assumed.
Before the merge, the stronger post’s unique section on slat spacing is copied across, internal links are pointed at it, and the weaker URL is removed from the sitemap. The pair’s combined impressions are the baseline the result is read against.
What to do, step by step.
In Search Console, list the queries for which two or more of your URLs earn substantial impressions.
For each pair, read both pages. If they answer different questions, sharpen each to its own intent instead of merging.
If they answer the same one, choose the page to keep by its history: links, conversions and the better answer.
Move the unique sections of the other page into it, then redirect the old URL permanently to the one you keep.
Update internal links, the sitemap and canonical tags to point to the kept URL, so every signal agrees.
Measure the kept URL’s impressions, position and clicks against the pair’s combined figures over the weeks that follow.
Mistakes to avoid.
Merging pages that serve different intents. A category page and a buying guide can rank for one query and both deserve to. Merging them loses the reader the second one served.
Redirecting without moving the content. A redirect passes a URL’s signals, but a reader who came for a section that was dropped finds nothing. Carry over what was unique.
Canonical tags as a shortcut. Pointing one live page’s canonical at another while both stay linked sends mixed signals. Google’s guidance warns against naming different canonicals for one page through different methods[1].
Rewriting a page that already performs. Google’s guide to traffic drops notes that small changes of position happen at any time and can reverse on their own, and recommends avoiding radical changes to a page that is already performing well[5]. Confirm the swap before acting on it.
How QueryRev treats it.
QueryRev finds the pattern in Search Console data by page and query, and proposes a Merge decision: which URL is kept, the redirect, and the sections to carry over, with a modeled value range. A person approves, and the result is verified afterwards.
Questions people ask.
How do I find keyword cannibalization?
Filter the Performance report in Search Console by a query and open its pages. Two of your URLs with substantial impressions for the same query, positions that trade places over weeks and a ranking URL that changes between days are the signs.
Should I always merge pages that compete?
No. Merge when the pages are interchangeable. When they serve different intents, make each one clearly about its own question and link between them.
Can a 301 redirect fix cannibalization on its own?
It moves the old URL’s signals to the new one, but it does not move the content. Carry over the sections readers came for first, then redirect, so the kept page answers everything the two pages answered.
Sources and citation.
Primary sources only: the search engines, the AI vendors, standards bodies and QueryRev’s own research. Each is cited in the text by its number.
How to specify a canonical URL with rel="canonical" and other methodsGoogle Search Central ·
Redirects and Google SearchGoogle Search Central ·
State of AI crawler access, October 2026QueryRev Research ·
Performance report (Search results): Overview and basic setupGoogle Search Console Help ·
Debugging drops in Google Search trafficGoogle Search Central ·
Cite this definition
A suggested citation. Its link goes to the definition, whose anchor does not change.
QueryRev Research. “Keyword Cannibalization.” QueryRev AI glossary, 18 October 2026. https://queryrev.com/glossary/keyword-cannibalization#definition
Related terms