Koogle d*EPiC SWYNme BITLYme EmailMarketingTipps


(Viewing source=hookedondata.org)

discussion Guidelines For Ab Testing

1) Have one key metric for your experiment. You can (and should!) monitor multiple metrics to make sure you don’t accidentally tank them, but you should have one as a goal. Revenue is probably the wrong metric to pick. It is likely a very skewed distribution which makes traditional statistics tests behave poorly. See my discussion in my A/B testing talk (around the 23-minute mark). I generally recommend proportion metrics. First, you often you care more about the number of people doing something than [...] [more]
hookedondata.org    Test
Page 1