View Single Post
Old 2009-02-17, 10:30 PM   #5
hony
Oh! I haven't changed since high school and suddenly I am uncool
 
Join Date: Dec 2004
Posts: 251
Pefect thing for A/B split testing. Provided the sponsor lets you have some kind of custom tracking then you just change between your two options at random until you have a large enough sample size.

When you have enough data, then you switch to the best option. Or better yet switch out the worst option and compare the best option with some new idea.
hony is offline   Reply With Quote