Yep. But it doesn't work very well... It assumes that all of the variables that you test are independent of each other, but they aren't. So your results are virtually worthless... For example, you may test four alternatives in each of four positions on a product page. And the best thing to stick at the top is a call to action. And the best on the left is a call to action. And the (etc). But you haven't tested four calls to action on the page - and even if you had, and it didn't work, the result could still be to use that option (since the best alternative in each position is to have a call to action - most of the time). The same is also true with advert text.