Keyword Cannibalisation

Keyword cannibalisation happens when two or more pages on the same website target the same query and the same search intent, so they compete with each other in search results. Instead of one strong page, you end up with several weaker ones, rankings that swap between URLs, and links and engagement split across pages that do the same job. As MLforSEO’s keyword mapping guide puts it, if the same keyword is mapped to two pages, you have planned cannibalisation before writing a word.

How it works

Search engines try to pick the single most useful page for a query. When your site offers two near-equivalent candidates, three things tend to happen:

• The ranking URL flips between pages over time, which makes performance hard to read and to improve.
• Internal links, backlinks and engagement are divided between the pages, so neither collects the full benefit.
• Google’s deduplication and site diversity systems limit how many similar results from one site it shows, so the weaker page often adds nothing.

Intent matters more than the words. MLforSEO’s search intent classification guide quotes Bernard Huang of Clearscope: there is no keyword cannibalisation, only search intent cannibalisation. Two pages sharing a keyword are only a problem when they serve the same need.

How it’s used in SEO and marketing

Detection usually combines three checks:

• Search Console queries: export page and query data (or use the Search Console API) and look for queries where several of your URLs earn impressions or swap positions.
• Keyword mapping: assign each keyword to one target URL. MLforSEO’s template flags the same keyword mapped to two URLs, and a ranking URL that is not your chosen target.
• Content audit: near-duplicate titles, headings and topics, which fuzzy matching or keyword clustering can surface quickly at scale.

Fixes depend on what the pages are for:

• Merge and 301: combine overlapping pages into the stronger one and permanently redirect the other. Google describes redirects as the strongest signal for consolidating duplicate URLs.
• Re-target: give the weaker page a distinct intent or angle and move the keyword to its secondary list.
• Internal links: point descriptive anchors at the page you want to rank, as MLforSEO’s internal linking workflow does after picking the best-ranking page for each query.
• Canonical tag: for near-identical variants (filters, parameters, print versions), a rel=”canonical” tells Google which version you prefer.

Example

In the MLforSEO keyword mapping sample data, “running shoes for wide feet” was mapped to both a wide-fit category page and a wide-feet buying guide, so the cannibalisation flag fired on both rows. The fix was to check the live results, see which page type matched the intent, assign the keyword to that page and keep it as a related secondary keyword on the other.

Limitations and things to watch

• Two of your pages ranking for one query is not always a problem. Google’s site diversity system generally shows no more than two listings from one site, and more when it judges them especially relevant. If both pages perform and serve different needs, leave them alone.
• Duplicate content is not a penalty. Google’s documentation says that without a canonical it will pick the best version itself; consolidation simply helps you control which page that is.
• A canonical tag is a hint for duplicates, so it may be ignored on two genuinely different pages. Merging or re-targeting is the cleaner fix there.
• Never delete a page outright: redirect it so links and bookmarks still land somewhere useful.

Learn it on MLforSEO

• Keyword Mapping: A Free Template and the Rules I Use
• Content Audit Template: Keep, Update, Merge or Prune Every Page
• Keyword Mapping Template with Intent and Cannibalisation Flags (Google Sheets)
• Search-Intent Classification with Machine Learning: The Approaches Explained