100 Split Tests

S1: INTRODUCTION: Split testing has been known to drive some pretty compelling results, but that all assumes one thing: that it’s done correctly. There are numerous aspects that you can look into for split testing, and they can get complicated. I’m here to make things clear for you. So, if you want to run an A/B test accurately and get long term results, you need to do these 3 things first: S2: Have a Clear Goal: is your goal to drive traffic? Is it clicks? Or even opt-ins? A clear result is determined by setting a clear goal at the start. An email subject line that wins at an open rate may be different than the winner for click-through rates. Define your goal first. S3: Only One Variable Is Tested: In A/B split tests, it is important that only one variable is tested. For example, you published two bonus pages. Bonus Page A at 10:00 A.M. on Monday with a green-colored buy button, and Bonus Page B, with an orange button and published on Monday 1 P.M. Bonus Page B wins in a click-through split test. In this case, because there are 2 variables, you won’t know what made Bonus Page A lose: the button color or timing? S4: Random Audience: If only your JV list sees Video Thumbnail A and only your buyers’ list sees Video Thumbnail B, can you straightaway use Thumbnail A if it wins an A/B test? You can, but it won’t be effective since different groups react differently. The best way to do split testing, apart from analyzing the effect of changing just one attribute, is to test your attributes with a completely random audience.