SERP Similarity Score

The SERP Similarity Score shows how many URLs overlap in the top 10 results of two search queries.

A keyword pair: two pages or one? The score gives you a number for that. It comes from tools, not from Google.

Where the term comes from

The term is an industry label without an official definition. It comes from keyword clustering. There, keywords are grouped by the overlap of their top 10 results, even if they share no wording. Wikipedia describes this method as a structure close to what a search engine dictates.

The vendor Thruuu offers its own tool for this and describes SERP similarity as comparing two keywords and their result pages. Because Thruuu sells the tool, this account is not neutral.

How the score is calculated

The principle is simple. A tool retrieves the top 10 results for both queries and counts the shared URLs. One example: if two queries share eight URLs, the score is 80 percent.

How many shared URLs are needed for grouping is up to each tool. Wikipedia calls this the adjustable “clustering level”, the minimum number of shared URLs. A low level produces a few large groups, a high one many small ones. Thruuu names four or more shared URLs, or 40 percent, as a rule of thumb.

How it differs from search intent and cannibalization

Search intent is the question behind a search. The score is an indirect measure of it. If many URLs appear in both result sets, that suggests Google serves both searches similarly. It is an inference, not a statement from Google.

Keyword cannibalization is the problem you want to avoid: two pages of your site compete for the same search. Thruuu recommends one page when overlap is high and separate pages when it is low. The score is thus a decision aid that can prevent cannibalization. It does not measure it. Whether it occurs on your site shows in your own rankings.

What this means for your website

Use the score when planning new pages. If it is high, combine the keywords on one page. If it is low, plan separate pages, each with its own focus.

Handle borderline cases by hand. At about 40 to 50 percent, Thruuu calls the decision an edge case and recommends checking the SERP and search intent yourself. Look at which page types rank: guides, shops, industry pages. That often says more than the number.

Limits of the score

The score depends on the moment of the query. Results change, also by location and device. Two tools can therefore return different values for the same pair. Compare values using the same tool and the same settings.

The score also looks only at the first ten URLs. It says nothing about the quality of the results. If the existing results are thin, two dedicated pages can make sense even at a high score, according to Thruuu. Anyone selling you a fixed threshold as law is treating a rule of thumb like a Google rule.

Frequently asked questions

Is the SERP Similarity Score a Google value?

No. Google does not calculate or publish such a value. It is a metric that SEO tools derive from search results. It is a hint of the same search intent, not proof of it.

From what value is one page enough for two keywords?

No standard defines this. Thruuu names 40 percent, meaning four shared URLs in the top 10, as a rule of thumb. At 80 to 100 percent, Thruuu calls one page a clear choice. Near the limit, you should look at the results yourself.

Does the score replace checking search intent?

No. It only replaces the counting. Whether two keywords mean the same question shows only in the results themselves, such as page types and content. Thin or weak results can justify two separate pages even at a high score.