Keyword Cannibalization: Find It and Fix It

Keyword cannibalization is when two or more pages on your own site target the same query and the same search intent, so they compete with each other instead of one page ranking well. Google does not penalize it as such. It is a cost you create by splitting links, clicks and attention across pages that should be one.

This guide is for site owners and marketers who suspect their pages are in each other’s way. It gives you a Search Console method to confirm true cannibalization, a test to separate it from harmless overlap, and a simple rule for choosing between merging, redirecting and differentiating.

On this page

• What keyword cannibalization is and is not
• Does keyword cannibalization hurt rankings?
• How to find keyword cannibalization in Search Console
• Is it true cannibalization or normal overlap?
• How do you fix keyword cannibalization?
• Merging and redirecting without losing links
• Differentiating pages that should stay
• Getting your keyword cannibalization audit started

What keyword cannibalization is and is not

Keyword cannibalization means several of your pages are built to answer the same question. Ahrefs defines it as multiple pages on a website targeting the same or similar keywords, and stresses that it is a real problem only when those pages hurt organic performance.

It is not the same as duplicate content. Duplicate content is the same text at several URLs, which a canonical tag can settle. Cannibalization is usually two different pages, written separately, that happen to want the same searcher. Different cause, different fix.

It is also not a manual action or a penalty. Nothing in Search Console will say “you are cannibalizing.” You infer it from the data, which is why a method matters.

Does keyword cannibalization hurt rankings?

It can, but only under specific conditions, and the loudest advice online overstates the harm. Ahrefs rejects the idea that sharing a keyword is automatically damaging: pages rank for many keywords, and pages with mixed intent often pull their own long-tail traffic. Semrush draws the same line, saying the test is whether pages serve different search intents, not whether they rank for the same term. It also notes that several strong rankings for a branded keyword are usually fine.

Google’s own documents explain both sides. Its ranking systems guide says the site diversity system generally will not show more than two web page listings from the same site in top results, though it may show more when it judges them especially relevant. So two of your pages can legitimately appear together. A third rarely will.

The real costs come from the same-intent case. Google’s guide to consolidating duplicate URLs lists why consolidation helps: you control which URL people see, signals such as links gather on one preferred URL, tracking gets simpler, and crawl time is not spent on near copies. Those are the costs when two pages chase one intent:

• Links and internal links split between two URLs, so neither looks as strong as one combined page would.
• Google may swap which URL it shows, so clicks and rankings jump around.
• The wrong page can win, such as a blog post outranking the service page that actually converts.

The diagram below shows the process we use to check for it.

Scroll the diagram sideways to see all of it.

How to find keyword cannibalization in Search Console

The most reliable free method is to look at one query and see which URLs Google showed for it. Google’s Performance report groups data by tabs such as Queries and Pages, so the same data can be sliced either way.

1. Open the Performance report for Search results. The default view covers the past three months, which is a fair first window.
2. On the Queries tab, sort by impressions. List the 10 to 20 queries that matter to your business, such as the ones your service pages target.
3. Click a query to filter the report to it. Switch to the Pages tab.
4. Write down every URL that appears, with its clicks, impressions and average position. Google defines average position as that of the topmost result from your site, so read each URL’s row on its own.
5. Change the date range and compare. If the leading URL changes from one period to the next, Google is switching between your pages.

The Semrush guide gives the same first step: filter by query and check whether more than one URL gets clicks and impressions. One practical limit: Google truncates report exports to 1,000 rows, so a large site should work query by query in the interface rather than export everything and hunt in a spreadsheet. Our Search Console guide covers the report setup in more depth.

You can also run a quick site:yourdomain.com "keyword" search. Ahrefs warns it returns many loosely related pages, so treat it as a hint and confirm in Search Console.

Is it true cannibalization or normal overlap?

True cannibalization needs the same intent plus a visible cost. Two URLs appearing for one query is only the first signal. Run each candidate pair through four checks.

The second check is the one most guides skip. Search the query yourself and look at what Google prefers. If every top result is a how-to guide and your two pages are both how-to guides, they are rivals. If the results mix a guide and a service page, search intent may explain why each of your pages holds a spot.

A useful rule of thumb, which is our judgement and not a Google threshold: if the second page earns almost no clicks for the query and shares its intent, treat it as a merge candidate. If both pages earn real clicks from different phrasings, leave them alone.

How do you fix keyword cannibalization?

Pick one page to own the query, then choose the lightest action that makes that true. Semrush describes the same first step: choose a preferred page per keyword, judging by backlinks, rankings and traffic. The decision tree below turns that into three outcomes.

Scroll the diagram sideways to see all of it.

A fourth outcome is to do nothing, which is correct more often than the checklists admit. If the pages have different intents and each earns its clicks, leave them.

Merging and redirecting without losing links

Merging is the right fix when two pages share an intent and one is clearly weaker. The steps are short, but each one protects something.

1. Choose the survivor. Keep the URL with the stronger links, rankings and traffic. Do not choose by which page you prefer.
2. Fold in what is worth keeping. Move any section, example or data point the weaker page does that the winner lacks. Ahrefs describes the result as one page that merges backlinks and internal links instead of spreading them.
3. Set a permanent redirect. Google says a permanent redirect is a signal that the target should be canonical, while temporary redirects (302, 303, 307) are not. Use a 301 or 308 when you are sure it will not be reverted. Our redirects and SEO guide covers chains and loops.
4. Update internal links. Point links at the survivor directly, not through the redirect, and drop the old URL from the sitemap.
5. Redirect to the closest match. Do not send the old URL to the homepage. A redirect only helps if the destination answers the same need.

Google ranks redirects as the strongest of its consolidation methods, ahead of rel="canonical" annotations and sitemap inclusion. The comparison below shows how the common fixes differ.

Scroll the diagram sideways to see all of it.

Two cautions on that table. Ahrefs calls noindex a “terrible way” to fix cannibalization and says canonicalizing suits duplicate content rather than keyword overlap. Semrush likewise notes that noindex does not consolidate ranking signals. Reach for those two only when the page truly has no search role.

Differentiating pages that should stay

Differentiating works when two pages serve genuinely different intents but their copy has drifted together. The fix is editorial, not technical.

• Rewrite the title and H1 of each page around its own intent, such as “how to” for the guide and the service name plus qualifiers for the service page.
• Change what each page promises in the first paragraph. A guide should define and explain, and a service page should describe the offer.
• Use internal links to tell the story. Link from the guide to the service page with descriptive anchor text, and stop linking to the guide with the service page’s target phrase.
• Check the keyword grouping. If you built pages from a list of near-identical terms, keyword clustering shows which terms really belong together.

After a differentiation edit, wait for Google to recrawl and compare the Pages tab again over a few weeks. If the same two URLs keep swapping, the intents were not as different as you thought, and merging is the next step.

Getting your keyword cannibalization audit started

Start with the five queries that bring you the most revenue, not the whole site. Run the query-by-page check, apply the four-test table, and decide each pair as differentiate, merge, canonical or leave alone. Writing the winner for each query into a simple keyword map prevents the same problem returning, as Semrush recommends, and checking that map before you publish anything new is the cheapest prevention there is.

If the pairs you find are mostly old, thin posts, the content pruning and consolidation checklist walks through merging and removing them safely. If you would rather have the whole site checked, including how overlap affects what AI assistants cite, our AI visibility audit covers it.