Tuesday, January 10, 2006

Split Test Or Die

If you are not split testing, you are crazy.

It is commonly known that the if your headline does not grab your visitors interest in a few seconds, she is lost. Gone forever. Bye bye.

So how on Earth is she going to buy anything from you, if she leaves your site?

She isn't!


So the better the headline, the more sales.

Now if you change your headline, how do you know if it is better? I mean, really better... (Or worse?)

How do you know it's not just the quality of traffic that changed?

Or if sales are just low because it is the holidays? (Oh, dreaded holidays.)

Here's how you do it: You split test. You do what is often called an A/B Split Test. In an A/B Split Test, you create two sales pages that look just the same, except one thing -- the thing you want to test, say, your headline.

Then you divert traffic evenly between the two pages.

That way, you keep everything the same (time of day, what day it is, if it's the holidays) -- except the ONE thing you are testing.

I use a free script called Split Test Buddy, and it works great.

So you have 200 visitors, say, and send 100 to page A, and 100 to page B.

3 people buy from page A, and only 1 from page B.

Now is A better than B?


In order to know, you have to do some stats. And because we don't want to be thinking about mathematics, only about marketing, we make the machine to the work.

So I've created a small A/B Split Test Calculator, that you can use for free. It has some instructions there for you, so don't by shy, try it out!