Sign up

Search Console analysis

How to find and fix keyword cannibalization with Search Console

Confirm keyword cannibalization in Google Search Console, tell harmful overlap from harmless overlap, and choose between merging, redirecting, re-targeting and relinking.

On this page

Part of How to analyze Google Search Console data

Keyword cannibalization is harmful when two or more of your pages split impressions for the same query and the ranking page keeps changing, so neither ranks as well as one strong page would. You confirm it in Search Console by opening the query and checking the Pages tab, then comparing over time. The fix is usually to pick one page and consolidate the rest into it, but not every overlap needs fixing. The glossary entry has the short definition; this page is about the work.

Harmful or harmless?

SignalProbably harmlessProbably harmful
Who gets the impressionsOne page gets almost all of themTwo pages each get a large share
Ranking page over timeStableSwaps week to week
PositionsBoth pages rank in the top 5 (you hold two results)Both sit between 8 and 30
IntentThe pages answer different intentsThe pages answer the same question

Finding it in Search Console

  1. 1

    Start from important queries

    Performance → Queries, last 3 months, non-brand, sorted by impressions. The Search Console analysis guide shows how to set up the brand filter and the comparison.

  2. 2

    Open one query

    Click it, then switch to the Pages tab. Note every URL with meaningful impressions.

  3. 3

    Check over time

    With the query filter still on, compare month by month. If the top page changes between periods, Google is undecided.

  4. 4

    Repeat from the page side

    Filter by one page, open Queries, and look for the queries where a sibling page also ranks.

Tip:This is slow for more than a handful of queries, because the interface shows one query at a time. Query-and-page data from the Search Console API shows every overlap at once; AnotherSEOGuru App's cannibalization detector runs on that data and lists each query with the pages competing for it.

Choosing the fix

SituationFixKeep in mind
Two pages say the same thingMerge into the stronger page and 301 redirect the otherMove the unique, useful parts; update internal links to the new URL
Pages serve different intents but overlap in wordingRe-target the weaker page to its own queryChange its title, H1 and opening to its own intent
One page should win, the other must stay (e.g. a product page)Relink: point internal links with the query as anchor to the winnerRemove the query from the other page's title and headings
Near-duplicate variants (filters, parameters, print versions)Canonical to the main URL, or noindexCanonical is a hint, not a command; check that Google follows it

Which page should win?

  • The one with more clicks over the last 3 months for the query family, not only for the one query.
  • The one with more external links pointing at it.
  • The one whose format matches what ranks: a list for list queries, a tool for tool queries.
  • When in doubt, the older URL, since it has the longer history.

After the fix

Record the date. Check the query's Pages tab after two to four weeks: one URL should hold most of the impressions and the position should improve. Compare the winning page's clicks with the same number of days before the change. If the old URL still appears, check that the redirect works and that your sitemap and internal links no longer point at it.

Cannibalization often shows up as a slow slide rather than a sudden drop, so it belongs in your content decay review too. Once one page owns the query, it is often a striking-distance quick win worth pushing into the top 3.

Is it bad to have two pages ranking for the same keyword?
Not if both rank near the top and serve different intents; holding two results can be an advantage. It is a problem when both rank lower than one page would and Google keeps switching between them.
Can I fix cannibalization with a canonical tag?
Only for near-duplicate pages. For two distinct articles, a canonical tag is often ignored. Merging and redirecting is more reliable.
How long does it take to recover after merging pages?
Usually a few weeks, once Google has recrawled both URLs and followed the redirect. Larger sites with infrequent crawling can take longer.
Is there a free keyword cannibalization checker?
Search Console itself, one query at a time, as described above. For the whole site at once you need query-and-page data from the API, which tools such as AnotherSEOGuru App read for you.

Do this on your own data

Connect Search Console once (read-only) and AnotherSEOGuru App runs this analysis for every site you manage, every week. Search Console features cost no credits on any plan.

Sign up freeOpen the live demo

Was this page helpful?

Sign up freeGet a quote