CUPED
A variance-reduction method that uses each visitor's behaviour before a test to make results more precise.
CUPED (Controlled-experiment Using Pre-Experiment Data) adjusts each visitor's outcome by what they did before the test started. Much of the noise in revenue comes from who the visitors are, and pre-test data explains part of it.
Removing that noise narrows the intervals without biasing the result, so a test can reach a decision with fewer visitors. It helps most when many visitors have history, such as returning customers.
A store whose visitors are mostly returning customers turns on CUPED for a revenue test; the 95% interval on revenue per visitor narrows noticeably for the same traffic.
Related terms
- Revenue per visitor (RPV)Total order revenue divided by visitors in a variant: a goal that combines how many visitors buy and how much they spend.
- Sample sizeHow many visitors a test needs per variant to detect the minimum detectable effect with the chosen confidence and power.
- Statistical powerThe chance that a test detects an effect of a given size when that effect is really there.