Hi Friends, I'm using website optimizer. I have one original landing page and I have designed different page with some new content and lay out to test. In that case which test should I take? AB or multivariate?
If you are testing 2 version of the page layout you should use A/B Split testing. Multivariate testing is for testing more than one alternative at the same time. For example, if you wanted to test different font sizes and colors for your heading, you would use multivariate testing to test a few sizes and colors at the same time to determine which combination works best. You usually need a bigger sample size (i.e., more traffic) for multivariate testing compared to A/B Split testing.
I'd be interestd in the results. If you trying to measure which variant converts into orders best I hope you have a lot of orders. We ran A/B testing on our site where we took 1,000 orders per day - but it is just not enough observations to statistically define if version A is better than the control, unless your control version is really rubbish - but any knowledgeable useability person could pioint that out wqithout A/B testing. It might impress your clients but I doubt you'll really be in a position to make sound statiscalally based decisions
Well, does it depend on products or what? If i had to check it with CPA affiliate marketing type, do u think it will be beneficial where different landing pages needs to be checked for conversion.
It is independent of the product. A/b or multivariate testing has nothing to do with what product you are selling on website. It is certainly recommended to use a/b or multivariate testing if you would like to make more $$ via your landing page. Thanks, Ravi