Keyword cannibalization is when a search engine constantly exchanges ranking between multiple pages or when multiple pages rank simultaneously for the same keyword but are similar enough to be consolidated.

These cases need fixing because Google will display only one page at a time or you could get more traffic by rolling content under one page.

Example: “seo case studies” 

For the keywords “seo case studies” and “seo case study”, we saw a group of blog posts constantly exchanging rankings in the SERPs.

Although each case study had a different URL, the phrase “case study” in each title and URL may have triggered cannibalization.

This meant that whenever we published a new case study, the old one was dropped from the SERPs and substituted with the new one.

Keyword diversification is when two or more pages rank for the same keyword simultaneously, and there is likely no benefit in consolidating.

There’s nothing to be fixed here. If Google ranked your site more than once, the system rewarded your content.

Example: “keyword rankings” 

For this keyword, we rank with:

• A glossary article ranking in the featured snippet.
• A landing page for a tool at #4 .

Do the pages swap rankings? No, they’ve been ranking simultaneously for over half a year.

Is there any benefit to consolidating? Unlikely. This way, we serve two kinds of intent: we cater to people who want to get a definition of the term as well as those who want to check their keyword rankings.

Example: “SEO audit” 

For this keyword, we rank with:

• A blog article at #4.
• A landing page for a tool at #7.

None of the pages swap rankings.

Consolidating them would make no sense because we would need to choose between catering to searchers who want a tool and those who want a guide.

Why rank once when you can rank twice, right?

Sure, if you rank two times in the top 10, you get traffic from both positions (as shown in traffic uplift in the table below). This may seem like an absolute gain in traffic, but it’s a bit tricky. In practice, the new traffic may not be worth it, and you’re also risking pushing down another valuable page.

Let me illustrate this. I took a bunch of keywords where we ranked with 2 URLs in the top ten (otherwise, we’d hardly get any traffic) and looked at the click distribution.

• Traffic uplift — the traffic we gained via ranking another page in top 10 (traffic to new page/traffic to old page).
• Traffic share of the old page — traffic share of the page that ranked first for the keyword.
• Traffic share of the new page — traffic share of the page that came later, i.e., the page that created diversification.

As you can see, pages that come later to the SERPs do not always get a bigger traffic share. Just by looking at this small sample, we can see that there are at least three possible scenarios:

• The new page can outrank the old page, either by jumping into a higher position than the original page or pushing the old page lower. More traffic will go to the new page, and in some cases (because that also depends on the CTR), you will gain significantly more traffic.
• The new page will outrank the old, but will get lower CTR, and less traffic. This is what happened in the case of “keyword rankings” — the featured snippet with a definition of a term ranked higher than a landing page with the free tool but got fewer clicks.
• The new page won’t outrank the old page or will rank in site links. Then your diversification efforts will add a few percent of new traffic at best.

Keep in mind that:

• All newly added pages ranked for their unique set of keywords, bringing in additional traffic not listed above.
• We’re only talking about the quantity of traffic here. How qualified that traffic was is a different story.

First off, it’s not something you can fully control. In our case, most of those instances were accidents — Google’s hard-coded propensity to rank some sites multiple times on the same SERP.

But let’s say you see this great opportunity to rank twice in the top 10. It’s definitely possible, but again, a bit tricky. Based on my observations:

Obviously, more popular keywords will have enough volume to distribute considerable traffic across the top 10. However, that traffic will largely depend on the CTRs, and these are different for each keyword.

The median search volume for cannibalized keywords is virtually the same as for our single rankings (50 vs. 40), so we can infer that search volume doesn’t play a determining role.

Theoretically, you could rank multiple pages either by aligning with search intents already present on the SERPs or creating something unique but extremely valuable.

If I had to answer this question having only our data, I’d say “follow the crowd”. As you will see in a bit, multiple rankings can happen both with keywords with a single dominant search intent as well as those with a more mixed intent but none of our pages broke outside of already existing types and formats of content.

We’d expect long-tail keywords to have fewer multiple ranking instances (be it cannibalization or diversification) because the intent would be more defined from the query.

To measure this, I compared word count distribution in keywords with single rankings and multiple rankings using boxplots.

As you can see, the boxplots are almost identical, so no correlation could be at play here.

However, an interesting takeaway here is that if we look at a histogram of word count frequency in the multiple URL data set, we see that most keywords occupied the 3–4 word count range. That indicates ranking multiple times for quite specific search queries is definitely possible.

Let’s sum up.

Multiple rankings in Google create two effects: cannibalization and diversification.

This means that classic SEO theory is wrong—you don’t need to go and fix each keyword with multiple rankings.

This also means that you can potentially take advantage of Google’s propensity for multiple rankings and can get more traffic out of a keyword. That said, I’d consider it a pro move, which may have a high opportunity cost compared to targeting some other keyword with just one page. That’s because:

• The additional rankings may not last long.
• You may “outrank yourself”, and the page you outrank will get fewer clicks.
• You may need more expensive types of content than what you already have. For instance, a free tool that serves the intent better than an article that already ranks.