Internet Marketing Digital_marketing_for_dummies | Page 337

Knowing How a Test Performed
For some tests, the data may overwhelmingly conclude that the variation you tested was a
winner or a dud. But if you have trouble determining how a variation performed, follow
these guidelines:
Your variation indicates it is successful and you should consider implementing it if
Your lift is statistically significant.
You show at least 100 conversions, or more, per variation.
Your variation indicates it is a failure and you shouldn’t consider implementing it if
Your loss is statistically significant.
You show fewer than 100 conversions per variation.
Your variation comes out null if
No statistically significant difference emerges.
The numbers have normalized.
The test ran for the entire test schedule.