Minimum Detectable Effect Calculator: The Smallest Lift Your Traffic Can Catch

A sample size calculator asks how many visitors you need. This one flips the question: given the traffic you can realistically get, what is the smallest improvement a test can actually catch? It matters because plenty of tests are doomed before they launch, chasing a lift the traffic was never large enough to detect.

The minimum detectable effect is the smallest relative change a test can tell apart from random noise. Low traffic does not stop a test from running; it just quietly raises the bar for what counts as a result, so anything smaller than the MDE will keep coming back inconclusive no matter how long you wait. If the figure below is larger than the lift you honestly expect, you are better off testing a bolder change, pooling more traffic, or accepting the test cannot settle the question.