View Single Post
Old 01-10-2011, 03:23 AM  
alf6300
Confirmed User
 
Industry Role:
Join Date: Sep 2007
Posts: 765
For a mainstream site we did extensive A/B testing on color combinations, painfully analyzed the results in detail, did careful validation, etc.

Conclusion: still not having a clue about what converts better, too many variables to consider.

My learning was that while test/analysis is important, overanalysis / paralysis is always a risk.
__________________
--
alf6300 is offline   Share thread on Digg Share thread on Twitter Share thread on Reddit Share thread on Facebook Reply With Quote