Sequential Testing
01 Definition
Sequential Testing meaning: Running experiments one after another, or evaluating results as data arrives over time.
Sequential testing means running experiments in a series rather than at once, or using statistical methods that allow checking results as data accumulates. In conversion work it often describes testing one change, applying the winner, then testing the next. Naive peeking at fixed-sample tests inflates false positives, so proper sequential methods adjust for repeated looks.
Also called
Why it matters
How Sequential Testing fits the work
Sequential testing suits sites that cannot run many variations at once, letting teams improve step by step and build on each confirmed result.
In context
Using sequential testing, the team first proved a new headline won, shipped it, and then began a separate test on the button color.
Practical note
If you check a standard fixed-sample test repeatedly and stop at the first win, you overstate significance. Use methods designed for continuous monitoring, or set the sample size in advance and wait.