If a domain ranks with several URLs for the same keyword, it has negative effects on the ranking. This is called keyword cannibalization.
This workflow shows you how to detect such cannibalisation in the Searchmetrics Suite and then provides tips on how to solve the problem.
Let's first navigate to the URL analysis in the Search Experience.
Search Experience > URL Performance > Overview
In the bottom of the page you will find a list of all URLs that rank with at least one of your project keywords at the time of the analysis.
Tip: If you see no or very few data here, it is probably because your project does not yet have enough relevant keywords. Take a look at all the workflows for keyword research to get some inspiration.
We recommend sorting this list by Traffic Index or Traffic Index Potential to filter out your most important URLs. To do this, click on the down arrow in the Traffic Index button. Keyword cannibalisation for these URLs could have a significant negative impact on the ranking and consequently on the overall traffic.
In our example, we now see that www.oracle.com is one of the most important URLs of our entire domain.
If you now click on a URL from the list, you will land directly in the URL details. Scroll down to the table at the bottom of the page. All keywords which rank with the selected URL are listed there. If not already displayed, add the column "Ranking cannibalisation" to the table. To do this, click the "+" in the last column and select the KPI mentioned (Tip: Working with elements).
As soon as keyword cannibalisation is detected, the number of own URLs that also rank for this keyword is displayed in this column. If an additional warning sign is shown in a row, this means that there is another internal URL with a better position on the same keyword.
Let's have a look at an example. We see that three of our URLs rank for the keyword "office libre dowlnoad". At least one of them is also better positioned in the SERPs.
To see the exact URLs, click on the number. A small pop-up will then open to show the URLs.
In our example, we see that even three internal URLs rank better for this keyword (positions: 1-3) than the analysed domain (position 4). We should now think about whether we can solve this.
a) Technical changes
If one of the URLs is an old or unimportant page, a 301 redirect should be implemented as soon as possible to direct users to the current page in the future and to symbolise Google which page is the right one.
Sometimes it makes sense to keep all URLs the same. In this case, it is recommended to use the attribute "noindex, follow" in the robots.txt of the weaker page to tell the Google crawler that the URL should be crawled but not indexed.
Another solution can be the use of canonical tags and the identification of the respective original page. This way, Google understands which page is the most important and can react accordingly. The use of canonical tags can be very helpful, especially in e-commerce.
b) Change content
If all URLs contain content that is important to the topic, combine them on the most important URL and redirect the remaining URLs to this URL via 301.
If possible, you should try to optimise a URL for a different/similar keyword. You may find a corresponding synonym and thus possibly even be able to cover the market even better.
Tip: You can discover similar keywords with Keyword Discovery. Take a look at this workflow to learn how it works: Keyword Research in the Keyword Research area.
Additionally, you could try to increase the relevance of the more important page by expanding the content based on the keyword and tailoring it more precisely. Also, by using more internal and trusted external links, the page can be ranked more relevant by Google.