Keyword Cannibalisation: How to Find and Fix Competing Pages

What is keyword cannibalisation?

Keyword cannibalisation is unwanted competition between pages on the same website that fulfil the same or very similar search intent.

Imagine a software company publishes:

/guides/seo-audit/
/blog/how-to-do-an-seo-audit/

Both pages are beginner guides explaining the same process. Their titles, headings and internal links target the same need. Google sometimes shows one URL for “SEO audit”, sometimes the other, while neither becomes the clear destination the business intended.

That is a plausible cannibalisation problem because the pages overlap in purpose, audience and outcome.

Now compare:

/services/seo-audits/
/academy/how-to-run-an-seo-audit/

The first sells a professional service. The second teaches someone to perform the work. They share a topic but serve different needs. Both can earn impressions for related queries without requiring consolidation.

The term is often written keyword cannibalization in US English and keyword cannibalisation in UK English. Both refer to the same SEO concept.

When are multiple ranking pages not a problem?

A website can legitimately have several pages about one subject. Search behaviour is varied, and one broad topic can contain different tasks, audiences and stages of a journey.

The pages satisfy different intent

A category, product, comparison and how-to guide can all discuss running shoes while helping people do different things:

/running-shoes/                 Browse the full category
/running-shoes/trail-pro/       Evaluate one product
/compare/trail-pro-vs-ridge/    Compare two products
/guides/choose-running-shoes/   Learn how to choose

The pages need clear titles, headings and internal routes, but shared subject matter does not make them duplicates.

The pages answer different levels of detail

A broad guide can introduce canonical URLs while a separate troubleshooting article addresses why Google selected a different canonical. The second page goes deeper into a distinct problem rather than repeating the introduction.

More than one result is useful

Google may show two pages from the same site when both are relevant. If each page attracts suitable queries and visitors reach the expected destination, there may be nothing to fix.

Performance is stable and intentional

One page may lead for commercial searches while another appears for informational variations. Review query groups and landing-page outcomes instead of expecting one URL to capture every phrase containing the topic.

One page is a navigational result

Branded searches can surface a homepage, login page, help article or product page together. These URLs are not necessarily competing for one content role.

What does keyword cannibalisation look like?

Cannibalisation rarely has one decisive symptom. Look for a consistent pattern supported by several signals.

The wrong page appears for the query

A glossary entry may rank instead of the detailed guide, or an old event page may appear instead of the current evergreen resource. The result fails to match the route the business wants searchers to take.

Two pages repeatedly exchange visibility

The leading URL for a query or close query group may alternate over time. This matters when the pages have the same purpose and the switching coincides with unstable or weak performance.

URL switching alone is not proof. Search results change for many reasons, including query wording, device, location, new content and competitors.

Impressions are divided between near-identical pages

Several pages may collect visibility for the same query group while none becomes a strong, complete resource. Their content, metadata and internal links may all make similar claims.

Search engines select an unexpected canonical

Very similar pages may be clustered, with Google selecting a different canonical from the one you intended. This is closer to duplication and canonicalisation than ordinary keyword overlap, but it can create the same practical symptom: the preferred page does not appear.

Internal links send mixed signals

Different articles may link “SEO audit guide” to different destinations. A navigation module may promote an older page while the sitemap and canonical signals favour a newer one.

The pages have become indistinguishable

Separate service, category or location pages sometimes begin with distinct aims but grow into copies as editors reuse sections. Similar titles, H1s, body content and calls to action make the intended difference hard to see.

How to find competing pages in Search Console

Search Console provides the performance evidence; it does not label cannibalisation automatically. Use the Search results Performance report to move from a query to the pages Google actually showed.

1. 01
   Choose a meaningful time range.

   Use a period long enough to reveal a pattern, then compare it with the preceding period when useful. Consider seasonality, migrations, redesigns, launches and major content changes. Short windows can exaggerate normal movement, especially for low-volume queries.

2. 02
   Start with an important query or query group.

   Open the Queries tab and select a query. For a subject with several natural variations, export the data or use a containing or regular-expression filter to group closely related searches.

   Do not combine phrases solely because they share one word. “Change business address” and “business address service” may represent different intentions.

3. 03
   Open the Pages tab.

   With the query filter applied, view the Pages tab to see which URLs received impressions and clicks for that query. Record clicks, impressions, click-through rate, average position, the date pattern for each URL, and country or device when they materially change the result.

   Focus on trends in impressions and clicks rather than treating small average-position changes as definitive.

4. 04
   Compare the URLs over time.

   Look for one page replacing another, visibility splitting after a new publication or an obsolete page continuing to receive the stronger signal.

   Example query-to-page comparison

   Query group	URL A role	URL B role	Pattern	Intended page
   seo audit	Beginner guide	Beginner guide	URLs alternate; clicks falling	URL A
   seo audit service	Service page	Beginner guide	Different variants	Keep both

5. 05
   Account for data limitations.

   Search Console omits some anonymised queries, and filtering or grouping can change totals. Page data may also be attributed to a canonical URL. Use the report to identify material patterns rather than expecting a complete keyword-ranking ledger.

6. 06
   Inspect the current search results.

   Review what Google currently shows for representative queries. Note the dominant result type and whether your pages actually match it. Treat the result as a snapshot: location, device, personalisation and timing can change what appears.

How to confirm the problem with crawl data

Once search data identifies a possible competing pair or group, compare what the website communicates.

Page purpose and search intent

State the job of each page in one sentence:

This page helps [audience] do [task] by providing [outcome].

If both sentences are effectively identical, competition is plausible. If one page helps someone learn and the other helps them buy, the distinction may be valid but poorly expressed.

Titles and H1 headings

Compare the title tags and H1s. Near-identical wording can make separate pages look interchangeable, particularly when the body content also overlaps.

URL A title: SEO Audit Guide: How to Audit Your Site
URL B title: How to Do an SEO Audit: Complete Guide

Rewriting these labels only helps if the pages genuinely serve different purposes. New wording cannot create a distinction that the content does not support.

Main content and format

Review the questions answered, examples, level of detail, audience, conversion action and content format. Calculate or compare similarity where useful, then make a human judgment about the pages' purpose.

Status, indexability and canonical targets

Confirm that each URL returns the expected status, allows indexing and declares the appropriate canonical. A supposedly competing page may actually be a duplicate URL that should already consolidate to another version.

Canonical tags are intended for duplicate or very similar pages. They are not a general tool for choosing a winner between two distinct articles.

Internal links and anchor text

Review which pages link to each URL, where those links appear and what their anchors say. Look for:

• different URLs receiving the same descriptive anchor;
• an older page receiving prominent navigation links;
• the preferred page buried several levels deep;
• links pointing through redirects;
• a hub linking to both pages without explaining the difference.

Sitemap and supporting signals

Check sitemap inclusion, breadcrumbs, structured data and related-content modules. The intended page should receive a consistent role across the site's systems.

How to choose the right treatment

The treatment follows the page relationship. Do not begin with the tool you want to use. Before adding more links, decide which page should lead; use a clear internal-linking strategy and check for overlapping content.

How to fix keyword cannibalisation

Keep both pages when the distinction is useful

If each page serves a separate intent, preserve both and make the difference obvious.

Align the:

• title and H1;
• introduction and primary task;
• depth and format of the content;
• call to action;
• internal anchor text;
• hub or category label.

For example, a service page should explain the engagement, deliverables and next step. A learning guide should teach the process without becoming a weaker copy of the sales page.

Differentiate pages that have drifted together

Return each page to a distinct audience or question. Remove repeated sections that do not support its purpose and add the information only that page should provide.

Do more than swap keywords. A meaningful distinction may involve changing the page type or format, adding unique evidence, tools or examples, narrowing one page to a specialist task, changing the conversion path, or linking the pages together with wording that explains their roles.

Merge overlapping pages

When two pages satisfy the same intent, choose the stronger destination or create a combined one.

1. Inventory the useful material and links on both pages.
2. Build one complete resource without copying every repeated section.
3. Publish it at the preferred, stable URL.
4. Permanently redirect the retired URL directly to the final page.
5. Update internal links, sitemap entries and other references.
6. Check that the destination remains indexable and self-canonical.

Choose the preferred URL based on page quality, relevance, link history, current performance and long-term fit—not simply which slug contains the keyword.

Redirect an obsolete page

Use a permanent redirect when the old page no longer needs to exist and a close replacement serves the same need. Do not redirect a specialist article to a broad homepage simply because both mention the topic.

Update the site's internal links after the redirect. Leaving them pointed at the retired URL hides the new structure behind an unnecessary hop.

Canonicalise true duplicates or close variants

Use rel="canonical" when duplicate or very similar URLs must remain accessible, such as tracking, print or product-route variants.

Make supporting signals consistent:

• link internally to the preferred URL;
• include the preferred URL in the sitemap;
• keep the canonical target working and indexable;
• use a self-referencing canonical on the preferred page.

Do not canonicalise two genuinely different articles merely because they target similar terms. A canonical may be ignored when the pages are not sufficiently alike, and it does not clarify their editorial purpose.

Change the internal targeting

If the content roles are sound but the wrong page receives the stronger site signals:

• update ambiguous anchors;
• link the preferred page from the relevant hub;
• remove outdated links to the wrong destination;
• link between the pages with a clear explanation of the difference;
• correct navigation, breadcrumbs and related-content rules.

Internal-link changes work best when the page content already supports the intended query and task.

Use noindex only for genuine exclusion

Noindex is appropriate when a page must remain accessible but should not appear in search, such as certain account, filter or internal utility pages.

It is usually not the first fix for two useful articles that overlap. Decide whether to differentiate or combine the content instead of hiding one page without addressing the user journey.

How to prevent competing pages

Map intent before publishing

Maintain a simple record of the primary job for important pages:

This is more useful than assigning one keyword to one URL without recording why the page exists.

Search the existing site first

Before commissioning a new article, review current pages, Search Console query data and the crawl. Update or expand a suitable resource when it already serves the same need.

Give hubs editorial control

Category pages should explain the difference between related resources and link to the right one for each task. Automated tag archives often connect pages without providing that guidance.

Review new pages after publication

Check which queries and URLs gain impressions over the following weeks. Early overlap may resolve as Google processes the content, so look for a persistent pattern before making another large change.

Audit during migrations and redesigns

New and old URL structures can coexist accidentally, redirects can be incomplete and navigation can continue to promote retired pages. Compare pre-launch URL lists, post-launch crawls and Search Console data.

Compare page signals in Clearcrawl

Clearcrawl provides the site evidence that ranking data alone cannot. It shows whether apparently competing pages use the same on-page labels, whether a canonical points elsewhere and which destination the internal link system actually promotes.

Begin with Search Console because a crawl does not know which queries generated impressions. Use the crawl to understand why the URLs may be competing and to verify the implementation after the decision.

How to verify the result

Check the live site

Recrawl the affected URLs and their linking pages. Confirm that:

• redirects point directly to the final destination;
• retired URLs no longer receive internal links or sitemap inclusion;
• retained pages have distinct titles, H1s and purposes;
• canonicals are valid and consistent;
• important internal anchors point to the intended page;
• the preferred page remains indexable and returns 200 OK.

Check Google's processed view

Use URL Inspection for important pages to compare declared and selected canonicals and see the last crawl information. Request indexing after a meaningful repair when the URL is important, then allow time for recrawling and reprocessing.

Monitor the query group

In the Search results Performance report, return to the same query or query-group filter and compare the Pages tab over a suitable period.

Look for:

• the intended URL receiving a clearer share of relevant impressions and clicks;
• retained pages separating into different query groups;
• improved combined performance after a merge;
• reduced visibility for retired URLs;
• more stable landing-page behaviour.

Do not judge the repair by whether only one URL ever appears. Different queries can still favour different pages. The aim is for each useful page to serve its intended need and for the site to present a coherent choice.

Common mistakes

Frequently asked questions