Keyword Cannibalisation
Last updated
Keyword cannibalisation occurs when multiple pages on the same site target the same primary search query. Instead of one strong page absorbing all the ranking signals for that query, signals are split across several pages. Google must pick which version to rank, and the result is often worse than what a single well-targeted page would achieve.
The term is also widely overstated. Multiple pages from one site appearing for the same query is frequently harmless: Google’s John Mueller has said that “pages aren’t duplicates just because they happen to appear in the same search results page”, and that three pages appearing for one query does not seem problematic “just because it’s ‘more than 1’”.1 Cannibalisation is not a penalty. The real issue is narrower: signals split across genuinely near-identical pages, or the wrong page ranking, where a single stronger page would do better. That is worth fixing; mere overlap between distinct, useful pages usually is not.
Why does cannibalisation damage rankings?
When two pages target the same query:
- Internal links pointing to either page divide PageRank that would otherwise go to one
- External links are split between URLs, weakening both
- Google’s quality signals (engagement, authority, relevance) are distributed rather than concentrated
- Google may alternate which page it shows, producing oscillating rankings that are difficult to track or improve
The effect is not always immediately obvious. Both pages may rank in positions 8 to 15 when a single page could rank in positions 3 to 5. The lost opportunity is invisible unless you compare current performance to what consolidation would likely achieve.
What are the symptoms of keyword cannibalisation?
Oscillating rankings: the most common symptom. A query shows your site in position 6 one day, then a different URL from the same site in position 11 the next. GSC will show impressions from multiple URLs for the same query.
Unexplained ranking instability: a page that should be well-placed for a query fluctuates erratically without any obvious external cause.
Rankings that stall despite site growth: as a site earns more links and authority, rankings for a specific query fail to improve. This can be a sign of split signals, but ranking stagnation has many other causes, so treat it as a prompt to investigate rather than a diagnosis.
How do you detect keyword cannibalisation?
Google Search Console Performance report: the primary tool. Filter the report to a specific query, then click the “Pages” tab. If more than one URL has received impressions for that query, you have at least partial cannibalisation. If the same query appears across three or four URLs with high frequency, the cannibalisation is active and likely harming rankings.
Site search with a query modifier: searching site:yourdomain.com "keyword phrase" in Google shows which pages Google associates with that phrase. Multiple results for the same phrase signal competition between your own pages. Several other operators narrow this further, and the full operator reference covers which ones Google still honours, since a number of the commonly cited ones have been retired.
Ahrefs or Semrush organic keywords report: sort by keyword, then identify cases where two or more pages from the same site share an identical or near-identical primary keyword. Semrush offers a dedicated cannibalisation report; Ahrefs surfaces the same overlap through its organic-keywords and Site Audit reports rather than a single named tool.
Crawl analysis: crawling the site and comparing title tags and H1s across pages can surface cases where two pages have near-identical targeting even if they are ranking for different queries currently.
Use the site: operator to spot keyword cannibalisation quickly
Search Google for site:yourdomain.com intitle:"key phrase" and you'll see every page on your site that has that phrase in the title tag. If more than one result appears, you likely have competing pages targeting the same term. This check takes under a minute and surfaces the most obvious cannibalisation cases before committing to a full audit. For queries where the conflict isn't in the title, use the ranking overlap reports in Ahrefs or Semrush to identify pages competing on the same keywords in search results.
What causes keyword cannibalisation?
Multiple articles on the same topic: common in content-heavy sites where older posts are not audited when new content is created. Three posts covering variations of the same question effectively compete for the same query.
Category and product/post pages targeting the same term: a category page and an individual guide both targeting “technical SEO checklist” will compete, particularly when neither is clearly stronger.
Location or variant pages using the same template content: location landing pages that differ only in city name will cannibalise each other for any non-location-specific query that appears in the shared template content.
CMS-generated duplicates: tags, categories, author archives, and date-based archives that reproduce the same content at different URLs.
How do you resolve keyword cannibalisation?
The right resolution depends on whether the competing pages serve genuinely different intents.
Consolidate (most common): if two pages target the same query with the same intent, merge them into one page. Move the best content from each into the surviving page, then 301 redirect the weaker page to the surviving one. The surviving page absorbs all inbound links from both URLs.
Differentiate: if the pages target the same query but serve subtly different intents or audiences, rewrite each to be clearly distinct. One might target an informational intent (“what is X”), the other a commercial intent (“best X tools”). The distinction must be genuine: superficial differences will not prevent cannibalisation.
Noindex or remove: if the weaker page has no search value and no inbound links worth preserving, removing it (returning a 404 or 410) is cleaner than a redirect. Use this for low-quality pages that should not exist, not for pages with any link equity.
Internal linking adjustment: in cases where cannibalisation is mild, ensuring internal links consistently point to the preferred URL can help Google identify which page should rank. This is a secondary fix, not a substitute for consolidation when two pages are actively competing.
How do you prevent keyword cannibalisation?
Cannibalisation is easier to prevent than to fix after it has accumulated across a large site.
Keyword mapping assigns one primary keyword to each page before content is created. When each page in the content plan has a distinct primary keyword, unintentional overlap is visible before any writing begins. Keyword mapping is covered in the keyword mapping cluster.
Content audit cadence: reviewing existing content every six to twelve months catches cannibalisation before it becomes a structural problem. New content should always be checked against existing pages targeting similar queries before it is published.
Pillar and cluster architecture: organising content into topic clusters with one pillar page per topic reduces cannibalisation by design. The pillar covers the broad topic; clusters cover specific subtopics. Overlaps between pillars and clusters are identified as part of the architecture rather than discovered after the fact.
Cannibalisation versus topical depth
Not every case of multiple pages on the same topic is cannibalisation. A site can have a pillar page on “technical SEO” and individual cluster pages on crawl budget, mobile SEO, and site migrations without those pages cannibalising each other, because each targets a distinct, specific query rather than the same broad one.
Cannibalisation occurs at the primary keyword level: two pages should not both have “technical SEO” as their primary keyword. They can both exist in a technical SEO topic cluster without competing, provided each has a clearly differentiated primary query.
