SEO Cannibalization: When Your Website Competes with Itself

the Keyword inscription made with 7 dice

You have two pages that discuss the same service. Both are well-maintained, both are indexed, and both are carefully written.

On Google, only one result appears at a time, and it changes from week to week. The Search Console report tells you that you’re ranked 11th for that search, but when you search manually, you find yourself in fifth place.

That number is correct, yet it tells a different story than it seems. It’s the average of two of your pages competing for the same search query, and the average of two distant positions indicates a point far from both of them.

It’s called SEO cannibalization. It’s one of the most common phenomena on websites that have been publishing for years, and it’s also one of the easiest to misinterpret.

Here’s how to spot it in your data, how to fix it, and what happens next—with a real-life example and actual numbers.

What Is SEO Cannibalization?

SEO cannibalization occurs when two or more pages on the same website compete for the same search query. The search engine selects only one of them to display, and that choice changes over time.

Visibility is spread across multiple URLs, each of which captures a portion of the signal, and the page you would have chosen ends up lagging behind.

This phenomenon almost always stems from the website’s growth. A sales landing page was published in 2019. In 2022, a blog post explored the same topic in greater depth.

In 2024, a guide tackles the subject once again, from a different angle. Every page has a good reason for being there.

The search engine, however, finds three documents that answer the same question and highlights only one of them—often the one you would have ranked third.

The overlap relates to the search intent behind the words. Two pages can use different vocabularies and answer the exact same question: to the search engine, they’re the same thing.

In marketing, it means something else

Those with a business background know this word in a different sense.

In marketing, cannibalization refers to a product that erodes the sales of another product from the same brand.

A beverage company launches a non-alcoholic version of one of its classic products and discovers that part of the new revenue comes from customers who previously bought the original: the market has grown less than the sales figures for the newly launched product suggest.

The logic behind the two meanings is the same. Two assets owned by the same owner are competing for the same demand.

Change the currency: in marketing, companies compete for revenue; in SEO, they compete for visibility in search results.
And in both cases, the aggregate figure—total sales, average ranking—masks the internal shift.

The Indicator Few People Notice: The Average Position

The average position that Search Console displays next to a query is the average of the positions of all the site’s URLs that appeared for that search, weighted by each URL’s impressions.

Consider a search query where your site appears on two pages. The first page ranks in position 6 and receives 1,000 impressions. The second page ranks in position 40 and receives 1,000 impressions. The calculation is (6 × 1,000 + 40 × 1,000) / 2,000 = 23. The report shows you 23, even though your pages are ranked 6th and 40th: position 23 is a midpoint between two very different realities.

This gives rise to two symmetrical misreadings.

  • The first: you see a decline where the page you’re interested in has remained stagnant, while the second one has been moving.
  • The second, more expensive option: you see an acceptable result where you actually have a strong page on the first page and a weak one on the fourth.

The solution is simple. A query’s ranking is always broken down by URL before drawing any conclusions.

With small sample sizes, the effect is even more pronounced: on a page that captures six searches, a new query that enters at position 30 shifts the average, while the actual ranking remains the same.

Our case: 11.7, which was actually 5.7

This case comes from the Factory Communication website, the first site on which we applied this analysis method.

The report indicated that the search term “SEO marketing consulting” had an average position of 11.7—a second-page ranking, thus requiring immediate action.

When broken down by URL, the picture changes.

URL Location Impression
/consulenza-seo-marketing/ 5,7 2.160
/consulente-seo/ 24,9 922
/consulenza-seo-per-aziende/ 82,0 3

The weighted average of the three rows is 11.5: the difference from the 11.7 in the report is due to rounding of the raw data.

The issue is something else entirely. The page designated for that search was already ranked at position 5.7, on the first full page, with the highest number of impressions.

The alarming figure came from the second row, a different page that appeared in the same search results at position 24.9 with 922 impressions.

If we had stopped at the aggregate data, we would have rewritten a page that was already working well. The real work lay elsewhere: determining which of the two was responsible for that research and giving the other its own scope.

How to Identify It in Your Data

All you need is Search Console and 15 minutes.

Open the Performance report, set a time period of at least three months, and sort the queries by impressions.

Choose a relevant query, click it to apply the filter, then go to the Pages tab.

You’ll see a list of the URLs your site displayed for that search, each with its own impressions and position. That’s where the overlay becomes visible.

Three signs to look for:

  • More than one URL with relevant impressions for the same query. Practical threshold: When the second page accounts for more than 20% of the impressions, the overlap is significant.
  • Changes in the URL over time. Compare two consecutive periods for the same query: if the top result changes, the search engine is wavering between your results.
  • An average position that moves while the content remains stationary. With small arrays, just one new query is enough to change the value.

A second check comes from the search engine itself: search for ” site:yourdomain.it ” followed by the keyword, and see how many of your pages answer that same question.

When you prefer a systematic review of the entire portfolio rather than a query-by-query approach, that is exactly the scope of an SEO audit.

How to Fix It

There are four options, and the choice depends on a single criterion: the search intent of the pages in question.

Consolidate

When two pages answer the same question and both contain valuable content, a single page is created that incorporates the best of both. The surviving page keeps the stronger slug, and the other is merged into it.

Narrow the scope

When each has its own audience and the overlap concerns only a portion of the text, both are rewritten with distinct purposes in mind: one addresses commercial needs, while the other addresses informational or in-depth needs.

This is the preferred option when a subpage generates a wide range of search queries, because keeping it means retaining that demand.

Redirect

A 301 redirect to the designated page is the right choice when the second page is brief and its content is already covered elsewhere.

Before activating it, make sure the relevant content is on the destination page: redirecting to a page that performs worse than the original is a step backward.

Exclude from the index

The noindex, follow tag applies to pages that serve users but make little sense in search results: service pages, archives, and taxonomies. The “follow” attribute preserves the flow of value through internal links.

The selection criteria are applied in three steps.

  1. Start with the intent: if it’s the same, then consolidate or redirect; if it’s different, then narrow the scope.
  2. Then consider the search volume: a page with a wide range of search queries is worth keeping and optimizing, even if it currently ranks low.
  3. Finally, it designates the main page, which is the one with the best ranking for the target query—rarely the most recent one.

Once the restructuring is complete, the focus shifts to strengthening the remaining page: this is the process we describe in “Ranking First on Google.”

What to Expect Next

A reasonable response time is six weeks.

Google rescans the pages, tracks redirects, and recalculates which document best answers that question.
The first two weeks often show significant fluctuations: this is the phase during which the search engine reindexes, and it’s best to let the content remain unchanged during this time.

The correct measurement is based on the same query set, the same number of days, and takes into account the seasonality of the industry.

Comparing July to May on a site that relies on spring searches leads to incorrect conclusions: first, you calculate how much overall demand has changed during that period; then, you analyze the individual page against that value.

There are three signs of success:

  • A single URL handles the query
  • The average position converges on that of the designated page
  • Clicks are starting to appear.

The third one comes last, and it appears when the article ranks in the top ten. On the fourth page, clicks remain a rare occurrence regardless of the quality of the headline.

Frequently asked questions

What is SEO cannibalization?

This is a situation in which two or more pages from the same website compete for the same search query. The search engine displays only one of them, and the choice varies over time. Visibility is split among multiple URLs, and the page you would have chosen receives only a portion of the traffic.

How can you recognize keyword cannibalization?

In Search Console, filter a query and open the Pages tab: if multiple URLs receive relevant impressions for the same search, there is overlap. A second clue is when the page displayed alternates between two consecutive periods.

Is any overlap between two pages a problem?

Overlap becomes a problem when it affects search intent. Two pages that include the same keyword but answer different questions can coexist just fine. However, when they answer the same question, the signal is split, and it’s best to take action.

Is it better to merge the pages or delete them?

Merging preserves the value of both and remains the preferred choice when the content is substantial. Deleting a page and redirecting it makes sense for thin pages that are already covered elsewhere. A page with a broad range of search queries, on the other hand, deserves to be kept and tailored to its own specific intent.

How long does it take to see results?

Six weeks is a realistic timeframe. The first two weeks show physiological fluctuations related to the reallocation. The assessment is based on the same set of queries and the same number of days, taking into account the seasonality of the sector.

Do you want to know how many of your pages are competing with each other?

Going through the search portfolio page by page provides a clear picture: which queries are served by the most URLs, which page is best suited to address each query, and which actions yield the highest return.

It’s the starting point for any well-executed reorganization.

Request an initial analysis.

If you are interested in SEO Strategies and Search Engine Optimization activities, you can also read these articles:

About this item
Share this article
If you need support, or want to understand how we can help your Company contact us now:
Would you like to receive information?
Fill out the form
This site is protected by reCAPTCHA and Google: Privacy Policy e Terms of Use.
Would you like to receive information?
Fill out the form
This site is protected by reCAPTCHA and Google: Privacy Policy e Terms of Use.