Glossary/Measurement & Attribution/Sample Ratio Mismatch (SRM)
Measurement & Attribution
Sample Ratio Mismatch (SRM)
When the actual traffic split between test groups differs significantly from the planned split, signaling a flaw that invalidates results.
By Brandon Mina·CEO, BrandPilot AI·Updated November 2026
Short answer
Sample ratio mismatch (SRM) occurs when the actual traffic split between test groups differs significantly from the planned split, such as 52/48 instead of 50/50 on a large sample. It signals a flaw in randomization, tracking or redirects that can bias results, so tests showing SRM are generally considered invalid until the cause is found.