How to Analyse and Fix Keyword Cannibalisation

Keyword and URL cannibalisation describes the, often negative, ranking effect that two competing pieces of content from the same domain can have in Google's rankings. By fixing this, you can improve domain visibility. In this tutorial we show you ways to find cannibalisation issues, and how to solve them in order to improve rankings.

Step-by-step example: E-commerce website canibalisation

Later in this article we’ll go into more detail on the types of cannibalisation and the multiple ways it can be assessed and fixed, but first let’s look at a common example where an online shop, and its category URLs, are competing against each other.

Select domain and view cannibalisation list

We’ve typed the domain asos.com into the search bar and selected the “Keywords” section in the main menu (left on desktop, burger menu on mobile screens.)

Now, select “Show Keyword Cannibalisation” 1 and you’ll be shown a list of all the keywords for which more than one URL from the domain is ranking. Click the “Search Volume” category title 2 to list the keywords with the highest average search volume.

The third column in the standard view 3 shows the number of URLs ranking. You can click on these numbers to show the URLs that are ranking.

Refine your canibalisation list

We advise you to ignore your brand keywords in this process. You should expect to be ranking for brand keywords with multiple URLs. We also advise to start with cannibalisation analysis in positions 6-10, where there are good opportunities.

If you would like to remove brand rankings, add the keyword filter “Does not contain the text”.

asos.com keyword rankings

In our example we’ve also added a position filter.

asos.com keyword rankings

In the final list above there are examples of the male vs female URLs, and of a sale directory that is competing with other URLs. What we are looking for are high-volume, unambiguous search terms for which a sub-category, or specific product, might also be ranking.

In some cases, the site structure might allow you to analyse a directory, such as ‘women’, in order to further refine the cases.

One example from the list above is the term ‘teddy coat black.’ The search term is well-defined but one of the two ranking URLs is less relevant than the other.

canibalisation example

We cover various scenarios in detail below but in this case, a re-working of the internal linking, with due consideration to other keywords that the URLs are ranking for, could solve the problem. A link to the “faux-fur-coats” page, in the relevant places, to the “teddy-coats” URL may be the answer.

How to fix keyword cannibalisation issues.

To fix cannibalisation issues, there are a number of options. The option you chose will depend on the type of issue you have and the type of resources you have available to fix it. For example, content and link changes can be done by editors whereas redirection may require a webserver administrator.

Internal Linking Structure

Internal linking (links that pass within the same domain) often gets relatively little attention but it’s a very powerful SEO tool. In-content links, with good anchor test, on strong pages, with few other links can tell Google, and users, a lot about how the site should be considered.

Read our SEO Basic tutorial on internal link optimisation for information about internal linking and how it can be used.

Considering internal linking may provide an easy solution when dealing with the problems of competing URLs

Merging Content

If you are generating content with the same purpose, ideally, you should concentrate all signals and effort into a single URL. This avoids having several different pages competing with each other. Only one of them earns all the authority. How can you manage this?

  • Canonical: The canonical rule tells Google which of the multiple pages contains the original content. This method isn’t always perfect, because Google reserves the right to respect the rule or to ignore it. It’s not possible to know for sure if Google’s algorithm is going to follow the instruction. This option can be useful in cases where we need to keep both URLs active and available, for example, so that they can be visited from various channels (email marketing, social media, etc.). The canonical link element would simply act as an indicator of preference for ranking and indexing the URL with the original content.
  • Redirection: By redirecting one URL to another, we are merging both pages into one piece of content. At this point, you should check the following:

    – Which page has achieved better ranking or better metrics.
    – Which page responds to the search intent more accurately.

    This way, we focus everything into one unique URL through a 301 redirect, and optimise all signals to point to the chosen page. We would keep all internal links pointing to the chosen URL instead of a ??redirected one??, and it would be included in the sitemap.

Deleting Content

Another possibility is to clean up duplicate content or content being used chaotically. We can erase content when there are many URLs cannibalising each other, but none of them gets good metrics (sessions, rankings, links).

  • Noindex: If we want to keep the content active for channels other than SEO, we can use “noindex” to keep this content out of Google’s index. We must be cautious in using “noindex” extensively, though, to avoid problems concerning internal authority distribution and crawl budget.
  • 404 or 410: There’s also the option of erasing the content entirely, and it’s feasible for pages, which haven’t generated any results, so we can “clean them up” to avoid cannibalisation and keeping indexed too much content on the same subject. Read more about 404 and 410 status codes.

What does “cannibalisation” mean in SEO terms?

Now that we’ve looked at a common example, let’s take a more detailed look at cannibalisation, how it manifests itself and ways to fix and prevent it.

Keyword cannibalisation is a common phenomenon which occurs when your own pages compete between each other in the search results, causing a problems with concepts such as:

  • Your organic CTR
  • Internal links
  • External links
  • Content
  • Traffic
  • Conversion

Cannibalisation sometimes happens due to lack of planning. It can occur as a result of different projects having an undefined structure, or, because the content strategy was not thoroughly planned, and instead is just being improvised. This leads to the production of similar pieces of content targeting the same keywords.

It can also happen with large websites that generate a lot of information and content, which occasionally end up repeating or duplicating content on different URLs tackling the same topic.

On the other hand, it sometimes comes up inadvertently, because a website has a very good authority and the search engine puts forward several pages from the same domain in the results, to accommodate ambiguous search intents.

There are ranking cases in Google’s Top 10 that might appear as ‘cannibalisation’ cases but are undoubtedly “positive” for the domain, as it would rank twice, or more, which in turn, increases the probability of getting better CTR. This can occur where a SERP has multiple intent or where it is brand-related.

You should also monitor cannibalisation issues stemming from pages deep in the architecture that might be getting other, strong, signals. This sends Google mixed ranking factors, causing the search engine to rank a different URL for the same keyword every week, for example.

Types of cannibalisation

Cannibalisation cases can vary greatly, but we’ve grouped them together into the following categories:

  • Several URLs ranking for one unique keyword
    • URLs have the same goal
    • URLs have a different goal
  • Two URLs experience a ranking “flip” for the same keyword
    • It could be due to the site’s page structure
    • It could be due to search intent issues
  • Duplicate titles or meta-descriptions, including the same keywords

As we’ve mentioned earlier, it’s possible to have positive cannibalisation, or “search result saturation“, which is done consciously using different content, to take up more space in the SERPs.

Negative cannibalisation, then, would be:

  • When there isn’t a correspondence between a ranking URL and the appropriate search intent. For example, when a blog post gets ranked instead of a product page.
  • When URLs with duplicate content are ranking due to structural deficiencies. For example, when filters are ranking instead of product categories.

Cannibalisation impact

Cannibalisation can affect any SEO project. Let’s round up the most important effects it can have:

  • Authority decline: when there are several URLs ranking for the same keyword, many metrics will end up splitting between the two, including CTR, traffic and authority, when they should congregate in one unique URL.
  • Internal and external links ranking signals are diluted: it’s not possible to strengthen ranking signals stemming from various links –either internal or external– for the same keyword or one that is fulfilling the same search intent. When assigning an anchor text to these URLs, which one would we use on each if they tackle the same topics?
  • Crawl budget split: keeping several pages focused on the same keyword will unnecessarily get them indexed and crawled twice, especially in large and medium-sized websites.
  • Conversion rate can break up and decrease: if there are various pages on the same subject, or responding to the same search intent, it’s very likely one of them is much better than the other ones. The fact that there may be other pages ranking for the same keyword could make the “good page” lose traction, and thus, sales.
  • Reduced visibility: Consider 2 URLs that rank at position 9 and 12. Focusing on one URL could boost it to better rankings.

How to detect cannibalisation?

To identify and detect which URLs and keywords are currently involved in the cannibalisation of our project, you can use various methods:

Manual search with search operators

Type-in the well known “site” operator to check which similar content for a specific query has already been indexed in Google.

It can be used in combination with other operators, such as:

intitle: to filter keywords contained within titles of the URLs.

inurl: to filter keywords contained within URL structures.

An example:

site:bbc.com + flapjacks
site:bbc.com + inurl:flapjacks
site:bbc.com + intitle:flapjacks

Google Search Console

Go to “Performance”, add your query, and switch to the “Pages” tab, to identify all URLs getting impressions and clicks through a query.

You can see that URLs ranking next to each other have the common denominator of a low CTR, which could be due to the snippet being under-optimised, or because it is competing among other URLs.

Note: Consider that the time range could include periods of content change on the website and include multiple countries for the same language.

Internal search

Occasionally, internal search can help us detect similar or duplicate results corresponding to the same query. However, keep in mind that this will depend on how the internal search engine is configured to return the results: match type, match query, upper case / lower case match, whether it takes into account relevance and meaning attributes, etc.

For example:

There are over 350 results for “cupboards” in IKEA UK’s internal search:

If there’s such a wide range of products of this type of furniture, then it might be worth considering this attribute for a standalone category or filter on the site, and thus, focus its search intent.

SISTRIX

SISTRIX provides three different ways to help you identify potential cannibalisation issues:

1 – Use ranking keyword history to identify which URLs from a specific website have been ranking, either from one week to another, or on a day-to-day basis.

If you look at the table, you can see which URLs are ranking every week, and if there’s evidence of “flipping”, it’s possible that there are several URLs fighting to get a spot in the results.

In the example above, you can already see that different content is interfering in the rankings of the “queen” keyword.

Taking the previous example of IKEA, potential cannibalisation can also be observed for the “cupboard” query.

2 – Use the “URL Changes” data

You can quickly identify which ranking URLs have changed for a particular keyword, between two dates of our choice.

Using the filter option, you can also analyse by keyword or URL groups. For example, all words containing “cupboard”, which will help to identify which URLs are fighting to rank for the same terms.

3 – Use the one-click Show Keyword Canibalization option.

From the “Keywords” section, we can use the filter called “Show Keyword Cannibalization“, and see the number of keywords for which there are more than one ranking URL. This information is displayed next to the “Position” column.

However, the results of this method can be interpreted in different ways:

  • The search intent is ambiguous and several URLs cater to this search intent.
  • The website has authority and ranks with several search results.
  • There are cannibalisation or structural issues.

How to reduce the chance of canibalisation

One way of reducing the chance of cannibalisation, which might suit a new site, or one being assessed for optimisation, is to work on the URL generation strategy.

Redesign your keyword plan

It’s possible that pages without a specific keyword intent assigned to them, all cover one particular aspect within a subject. All pages then end up competing between each other.

Example: when working with more generic terms, you should consider a content guide, or putting more effort into optimising the category gathering all the posts, while each article appropriately targets its individual keywords

On the other hand, you need to also keep in mind the existing demand for content if you’re focusing on long tail terms, for example, or that most searches are centred on the more generic term. In this particular case, if we analyse the SERP, it’s a mixed KNOW and a DO type search.

Change key SEO elements

Now that we’ve cleared up the previous point, each piece of content should be conveniently optimised. To help you with that you can use Keyword Discovery for a swift analysis. See which keywords are more in demand, and use them to optimise your content accordingly, or create new pieces, each focused on its own subject.

We can also keep in mind who is currently leading in the SERPs:

And, of course, see which modules are appearing in the results to determine which format you should use, too.

Approximately 38% of keywords return video and image results. Knowing this, and by analysing the SERPs for user intent we can create an optimisation table, as shown below:

TypeTarget keywordIntentWorking Title
CategoryglampingMixedAll you need to know about glamping
Postglamping with hot tubKNOWWhere are the best glamping sites with hot tubs?
Postglamping tentsDOThe glamping tent buyers guide
Postglamping podsDOFind a Glamping Pod that Suits You
Postglamping scotlandDOGlamping in Scotland - The best of the best.

We could add meta description, URL and H1 to this table. When the table is complete, it is possible to plan an internal linking strategy, use the appropriate anchor text in each case and boost level 1 – namely, a category– by, for example, using breadcrumbs.

We hope that from now on you will be able to contain and solve all your SEO cannibalisation problems!