What is keyword cannibalization?

Keyword cannibalisation is when two or more pages on your site compete for the same search intent, so Google has to pick between them. Common symptoms include ranking positions that swap between URLs week to week, a page that plateaus just outside the top ten, and click-through rates lower than the position should earn.

It’s all about intent

The common misreading is that using the same keyword on two pages causes a problem. It does not. A blog post about choosing a supplier and a service page selling that service both mention the term and both deserve to exist, because they answer different questions.

The problem is two pages that answer the same question. Three near-identical location pages. A 2023 guide and a 2025 guide covering the same ground. A service page and a landing page built for a campaign that nobody retired. Google picks one, often the wrong one, and changes its mind.

Confirming it with your own data

Search Console settles this faster than any tool. Open the performance report, filter to a single query, then switch to the pages tab. One URL is normal. Two or more URLs picking up impressions for that query over the same period is the signal worth investigating.

Then check the history. If the ranking URL for that query has changed more than twice in three months, Google is genuinely unsure. If one URL takes almost all the impressions and the other picks up a handful, that is usually a harmless overlap rather than a fight.

A site search operator helps as a second check. Search site:yourdomain.com plus the query and look at what comes back in the first few results. If you see three pages that could each plausibly be the answer, so does Google.

Fixing it

There are four honest options and the right one depends on why the duplication exists.

• Merge. Two thin pages covering the same ground become one stronger page. Move the useful content across, then 301 the weaker URL to the survivor.
• Differentiate. Both pages deserve to exist, but the angles are blurred. Rewrite titles, headings and opening paragraphs so each one clearly owns a different question.
• Consolidate links. Sometimes the content is fine and the internal linking is the problem. Point your internal links, in body copy and navigation, at the page you want ranking.
• Delete and redirect. Old campaign pages and expired offers with no traffic and no links should go. Redirect to the closest relevant page rather than leaving a 404.

Canonical tags are a fifth option, but a weaker one. They work when two URLs genuinely serve the same content, such as parameter variants. Using a canonical to paper over two different pages competing for the same intent just hides the decision from you.

Catching it before it starts

Most cannibalisation is created by process failure, not by writing. Nobody checked whether the topic already existed, so a new post went live on the ground an old one held. A keyword map that assigns one primary query to one URL, kept in a spreadsheet everyone actually opens, prevents nearly all of it.

We run that mapping as part of our data and reporting work, alongside a quarterly check for query overlaps in Search Console. It takes an hour and usually finds two or three pages quietly undermining each other.

Common questions

Does cannibalisation carry a Google penalty?

There isn’t a penalty. The cost is split signals: links, engagement and relevance spread across two URLs instead of concentrated in one, so neither ranks as well as a single page would.

Can a blog post and a service page target the same term?

You can, as long as they serve different intent. The post should answer a question; the service page should sell. If the post starts pitching your service, they are now competing.

How long until a fix shows in rankings?

In most cases, it takes two to six weeks once the redirect or rewrite is crawled. Larger consolidations on big sites take longer because Google has more URLs to reprocess.

For a short definition and related terms, see keyword cannibalisation in our SEO glossary.