A/B Testing Pitfalls

Using too large a sample size, which wastes resources Repeatedly checking results before reaching the required sample size (peeking), which inflates the false positive rate Running the test for exactly one week regardless of traffic A/B tests always give correct results as long as two groups are compared