
Why Do Multi-Variable Testing?
04-Feb-2012
Wayne Sharer
Multi-Varialble testing is a more advanced way to split test, though it’s not as difficult at the name sounds. It makes testing much faster, so therefore it is something you should know about to speed up your profit gains.
Multi-Variable Testing
One level up from A/B split testing is multi-variable split testing. Just as the name suggests, multi-variable split testing involves split testing multiple variables simultaneously.
The main benefit to this is its ability to find the best possible combination of variables.
Here’s an example of what this means:
Take for example testing a black headline versus a blue headline. The blue headline converts better. You toss out the black headline.
Next, you test a white background or a light grey background. The white background converts better. You toss out the light grey background.
Unfortunately, one combination that’s never tested in this case is the black headline on grey background combo. It never gets tested because the black headline was discarded after the blue headline converted better.
If you just did A/B split testing, there are many combinations of variables that you’d miss out on, which could result in a lot of lost profit.
The second reason marketers prefer multi-variable testing is simply because of how easy they are to setup. Setting up A/B split test after A/B split test is extremely time consuming, especially if you go through tests quickly. Being able to just setup one test and let it sit is often a big time relief.
Full Factorial vs. Fractional Factorial
There are two different ways to do multi-variable split testing.
The first way is called full factorial split testing. This means every single combination of variables is tested to its fullest extent. Doing this requires the most traffic but also gives the most accurate result.
The second method is called fractional factorial split testing. Instead of testing all variables thoroughly, this method involves testing a select set of variables to mathematically figure out what the best combination will be.
This will allow you to conduct multi-variable split testing without nearly as much traffic. Unfortunately, it does come at the cost of some degree of accuracy.
Which method you use depends mostly on how much traffic your web page gets and how accurate you need your data to be.
You now understand the basics of what split testing is and why it’s important. You also understand the various different methods available for split testing.
The next piece is to understand the crucial mindset that you must have to be successful in split testing.
International speaker, author, and entrepreneur. Retired navy officer, former commanding officer. Over 35 years of leading, coaching, mentoring, and speaking.
Read more