What Is Keyword Cannibalization and How Do You Fix It
Keyword cannibalization is when two or more pages on the same site target the same search and the same intent and the overlap costs performance: the wrong page ranks, or clicks, links, and relevance split so that neither page does as well as one consolidated page would. Two URLs appearing for one query is a reason to look, not proof of harm. You fix it by deciding what each page is for: merge pages that serve the same intent, differentiate pages that serve different intents, and use canonical tags or noindex only where a page must stay available.
Google’s guide to ranking systems explains why overlap costs space: “Our site diversity system works so that we generally won’t show more than two web page listings from the same site in our top results.” Pages from one site that answer the same search compete for those listings.
How It Happens
Cannibalization comes from ordinary publishing without a shared map:
- Different writers or teams cover the same topic at different times.
- Titles differ but intent does not, such as “best running shoes for flat feet” and “running shoes for flat feet buying guide.”
- An online store indexes both a filtered URL and a category page for the same products.
- A new page is published to refresh a topic instead of updating the existing one.
- Templated city or service pages differ only in a place name.
Diagnosing It
The Search Console Performance report is the main tool. Filter by a query, then open the Pages tab. If two or more of your URLs receive impressions for the same query, with middling average positions, look closer. Then check the reverse: filter by each page and compare the queries it earns.
| Pattern in the data | What it suggests |
|---|---|
| Several of your URLs receive impressions for one query, with shifting positions over weeks | No single page is clearly the site's answer; check whether combined clicks suffer before acting |
| One page ranks well and a near-duplicate ranks poorly for the same query | The weaker page may add nothing distinct; compare the two before merging |
| Two pages both sit beyond the first page for a query you expect to win | Links and relevance may be split between them |
| Overlapping pages with many impressions and low positions | The pages may answer the query less clearly than competitors |
Two quicker checks help on larger sites. A site: search for a phrase, such as site:example.com "flat feet running shoes", lists your own pages on that topic. A crawl export of URLs, titles, and H1s, grouped by main topic, shows overlaps across the whole site.
Choosing the Fix
- Merge when the pages serve the same intent. Combine the best of both into one complete page and redirect the other URL to it with a permanent redirect. Google’s redirect documentation explains that Google treats a permanent redirect as a signal that the target should be canonical, and the old URL stops competing. Before merging, list the queries, links, and conversions each URL earns, so the merged page keeps what matters, and afterward check that the combined performance holds.
- Differentiate when the intents differ. A “how to choose running shoes” guide and a “shop running shoes” category page can both stand. Rewrite titles, headings, and content so each page’s purpose is unmistakable, and link between them with descriptive anchor text.
- Use a canonical tag when both URLs must stay available, such as sorted or tracking versions of a page. Google’s canonicalization documentation ranks
rel="canonical"annotations among the strong canonicalization signals, one signal weighed with redirects, sitemaps, and internal links, so conflicting signals can lead Google to choose another URL. Where merging is possible, a redirect is the clearer fix. - Use
noindexsparingly, for pages users need but searchers do not, such as some filtered views. It keeps them out of the index, but not out of the crawl: Google’s crawl budget guide notes that Google “will still request, but then drop the page when it sees a noindex meta tag,” which spends crawling time. Neither tool merges pages the way a redirect does:noindexremoves a page from the index without passing anything on, and a canonical tag asks Google to consolidate only pages that are duplicates.
Why Another Page Is Not the Answer
When a page underperforms, publishing a new one on the same topic adds a third competitor to the two that already overlap. Before commissioning new content, check whether a page on the topic exists and should be updated, expanded, or restructured instead. Give each URL a clear job; where two pages do the same job and neither adds anything the other lacks, those two are the candidates to merge.
Preventing It
Keep a content inventory that maps each URL to its main query and intent, and check it before anything new is written. For sites that generate pages at scale, by city, product variant, or filter combination, decide in advance which page types are meant to rank and which exist only for navigation, and set canonical or noindex rules for the latter from the start.
To begin, take the last 90 days of Performance data, sort queries by impressions, and look for any query that is split across more than one of your URLs.