After all, if 12 folks click on on a button in e mail A and sixteen individuals accomplish that in choice B, you can’t actually tell which button performs better. Make your sample measurement large enough to get statistically important results. Most individuals skip e-mail A/B testing in