SERP overlap measures how similar the search results are for two different keywords, usually by counting shared URLs in the top ten. If “employee onboarding software” and “onboarding platform” share most of their top results, Google treats them as the same need. SERP overlap is the main evidence used in keyword clustering and one of the most practical ways to decide whether a keyword gets its own page.
How it is measured
The simplest version counts shared URLs between the two top-ten lists. Some tools weight matches higher when they appear near the top, or compare domains instead of URLs for a looser match. Pick one method and use it consistently, since the threshold you set depends on it.
overlap = shared URLs in both top-10 lists
Example:
keyword A top 10: a, b, c, d, e, f, g, h, i, j
keyword B top 10: a, c, d, k, l, m, f, n, o, p
shared: a, c, d, f -> overlap = 4Interpreting it
High overlap, often taken as four or more shared URLs in the top ten, suggests one page can rank for both. Little or no overlap suggests different intent or different page types, so separate pages are usually needed. The middle ground needs judgment: look at whether the shared results are the strong ones near the top, and whether the non-shared results are a different page type.
These thresholds are rules of thumb used by practitioners, not values published by Google.
Uses beyond clustering
Overlap checks help diagnose cannibalization: if two of your pages target keywords with high overlap, they are probably competing. They also help with intent shift: if the overlap between a keyword and its old cluster falls over time, the keyword may have changed meaning.
Example
One page or two?
Imagine Tidewhistle, a fictional knowledge base tool, wonders whether “internal wiki software” and “company knowledge base” need separate pages. Comparing results shows only one shared URL.
The wiki results are mostly product lists, while the knowledge base results include guides on building one. Tidewhistle creates two pages and links them to each other.
Common mistakes
Comparing results from different locations
Overlap is only meaningful when both result sets come from the same country, language and device.
Questions
Usually, but check business value too. If the second keyword is minor, it may not justify a page even with different results.
