6 Hot Tips For Successful Split Testing

How do you know if your marketing works if you’re not testing?

Do you casually glance at your website traffic stats, hold a finger in the air and say …yes that headline increased our website sales? Or do you scientifically test one headline against another and synchronize this with the number of orders you get?

To help make sure you’re on the right track here are 6 “hot” tips for successful split testing…

1. Test at least two versions of whatever it is you’re testing, this would be the contol and the variation

The control version is the one that’s getting you the best results so far, the variation is the version you’ve created that you want to have beat the control.

2. Equally test one version against another

You’re looking for each version you’re testing to be exposed to the same environmental factors. In the case of your website this would be making sure each version gets the same amount of website traffic form the same source.

3. Measure the actions of each split test

Now you have equal amounts of traffic going to each version you’re testing how will you determine which is more effective? This is where you need to state clearly the objective you’re aiming for, it might be more subscribers or more sales. Then you need to measure this for each version you’re testing.

4. A big enough sample of results.

It’s no use testing 10 visitors to each versions of you’re testing, I promise this won’t tell you anything of any value. On the other hand if you looked at 300 visitors for each test variation you’ll begin to see a more representative picture as to which version is the winner (which then becomes your control in future tests).

5. You need patience

Testing takes time, if your website doesn’t get much traffic on a regular basis it’ll take longer for you to get the results you’re looking for. On the other hand a high traffic site will get your website testing moving along much quicker, leading you to a high converting website in a fraction of the time.

If traffic is something you need more of recently I launched a traffic course called “Flood Your Site With Traffic” if you don’t have this yet click here to get yourself a copy today.

6. You need to understand the results

This one needs more investigation because sometimes you’ll find yourself in a situation where one version A of a test out performs version B by 5%, yet version B communicates your websites values more effectively. So which version should you choose? Over time you’ll become more experienced at testing and your “gut feeling” will play a significant role in deciding the winners of split test where the results are close.

To get you started with your split testing I recommend you take a close look at using Google’s excellent website optimizer.

Get started with your split testing today:

Google Website Optimizer…
http://services.google.com/websiteoptimizer

Split testing doesn’t stop with regular websites either… you can also split test the headlines and the content of your blog posts using this special WordPress plug-in…
http://websiteoptimizer.contentrobot.com

Here’s a useful online tool that helps you to calculate the effectiveness of your testing…
http://www.splittester.com/index.php

There’s one last thing you’ll come to realize once you’ve been split testing your website for a while… even your best testing is uncertain at best.

Why? Because of Heisenberg’s Uncertainty Principal which states that we can’t be certain of anything because the very act of looking changes its nature. This is why as Mark Joyner says in his excellent book Mind Control Marketing… “Drawing broad conclusions from your testing can be hazardous to your financial health”.

Tagged as: google website optimizer, headlines, internet marketing, marketing testing, scientific marketing, split testing