Question about A/B testing on ppc?

[kivd1]

It would be really difficult to explain my question theoretically so Im just gonna give an example.
Say I want to test the effectiveness of ppc>youtube video>affiliate page against ppc>affiliate page
but i also wanted to test the effectiveness of two different ad titles.
Should I test the youtube thing first and see if thats more effective and then test the ad titles on the more effective method? Or should I test both ad titles for both methods?
It seems like a simple thing when only talking about two things, but once you start testing all kinds of different variables in the campaign, testing out each method with each variation could be a very costly and difficult process

Thanks

 

[Endire]

Kivd1,

If you are doing the a/b test, I would recommend only changing one variable. If you are changing more than one variable (a multivariate test) then you are right, it becomes more difficult to track and analyze data.

Heres a good post on the differences between the two methods,

http://blog.hubspot.com/blog/tabid/...rence-Between-A-B-and-Multivariate-Tests.aspx

Best,

Shawn

 

[ouchthathurts]

I'm not sure if this will help you but its worth a shot, check out Google experiments they have split testing tools

 

[lluks]

use prosper202
google it

 

[Casey101]

use prosper202

I second this. Multivariate tests will take significantly more traffic than simple A/B tests. Also don't forget about statistical significance as it can cause problems for you if you ignore it. Good luck!