Just a quick question about the A/B test you ran: It looks like the structure of the signup page changed as well as the background. Don't you feel taht could've affected your results? I have little experience with A/B testing myself but have usually read that it should only be done by changing one aspect that you want to track. Was your A/B test different?
We are changing structure as well; if you independently look at A/B while also testing X/Y (of a different test), the results can be statistically misleading...in these cases we look at AX v. AY v. BX v. BY
3 comments
[ 3.2 ms ] story [ 18.9 ms ] thread