Don't discount the potential impact of randomness on your results, especially when the difference isn't statistically significant. If you find yourself micotesting every minor difference and making big changes on the smallest of differences, give this a thought.
Sign up to get updates from us
