Is Split Testing Obsolete?

Is Split Testing Obsolete?

I trust that your testing is going well.

Mostly the work that we are doing now is landing page optimization which we are calling the battleground for online profitability

Over the past few weeks we have developed a listing over 100 different items that can be tested on a landing page. At first, the biggest concern was around the number of success factors that the system required. Most could not think of 11 items to test.

Now with over 100 there should not be a shortage of things to test.

In fact, with that large a list it would take a very long time or a lot of traffic to get results if we were using the one-factor-at-a-time standard split testing method.

We are testing language, images, color, offer, call to action, popups/popunders, headline, pre-headling, sub headline, opening paragraph, closing, signatures, buttons, color of buttons, etc

But as you know with the MR2 you can have results in a few weeks. And keep improving and testing as long as it makes
sense to do so.

Think about it, before you could only test 1 item at a time and now you can effectively test 11 plus their interactions with each other.

2 to the 11th power is 2048 tests for your landing page. It would take your forever to do this the old way.

Here's a Direct Marketing Association article written in March 2004, I think that you will find it interesting.

That Voodoo We Do - Marketers Are
Embracing Statistical Design of Experiments By Richard Burnham, Founder,
Publication Coordination

http://www.the-dma.org/cgi/dispnewsstand?article=1958