Split Testing: the Scientific Method for Marketers

Today I’m combining science with marketing, specifically “Split Testing”. 

That gave me an excuse to post an image of a delicious banana split! Plus I’ll be mentioning “monkeys” soon and monkeys love bananas!

Marketing can be seen as an art form that works with humans on an emotional level. But there are parts of marketing that are more logical, more scientific.

The marketing split test is one way of breaking things down to the simplest level of making one change and testing which version performs better.

This is like posing a hypothesis in the scientific method. A hypothesis has to be falsifiable. 

By testing, there has to be a way the hypothesis could be proven false.

The classic example of this is making the hypothesis that, “All swans are white.”

If we came across even one swan that was some other color, say black or pink, then the hypothesis is false!

For marketing split-tests, we are making a hypothesis that this one small change to our marketing will perform better than the other version.

You can split-test things like the color of a website. You’ve had a red-and-yellow website theme. Now you create a blue-and-green theme. 

You then drive traffic to your site and have 50% go to the old version and the other 50% to the new version.

You can track the traffic and see which color scheme results in better stats like visits, clicks, opt-ins, sales, and profit.

This is a powerful way to test changes and continually improve your marketing performance.