April 15, 2014 By BJ Cook,

If you’re familiar with website optimization, then you've probably heard of A/B testing and multivariate testing. But, you may not be familiar with which type of testing you are using and how they only vary slightly from one another in uses.

A/B Testing

  • What is A/B Testing?

    • mostly referred to as split testing, where two versions of a webpage are compared to one another. Site visitors are either sent to version A or version B of the site and the testing is tracked by how they interact with the page.

  • How should you use A/B testing?

    • One of the most common uses is to test two completely different page designs. Or when one element and one element only is up for debate (i.e. CTA button) But, it is important to know that you can only track the design as a whole rather than multiple individual elements.

  • What are advantages to using A/B testing?

    • Since you are only testing a site design as a whole or an individual element the data isn’t overwhelming and can be achieved quickly. This is a great way to introduce website optimization to an unfamiliar or hesitant crowd. As your results are gathered in a shorter period of time.

Multivariate Testing

  • What is Multivariate testing?

    • Similar to A/B testing, but compares more than one element. Multivariate testing compares multiple elements, how they interact with each other and how the combination of elements make a version more successful (size, color, placement, etc.).

  • How should you use multivariate testing?

    • The most common use of multivariate testing is on a page that has multiple elements up for debate. If it’s headers, footers, sign ups, etc. or a combination of all three - multivariate testing will help you gather data on what design is going to be more successful for your consumer, based on the combination of modifications you test. Note: the more elements that need testing the longer it’s going to to take to collect the results.

  • What are advantages of using multivariate testing?

    • Once design data is determined and you know what elements your consumers prefer, you will be able to continue to design landing pages and campaigns based off of the data.

In short A/B testing and multivariate testing are the same type of test, just multivariate tests are done on a combination of elements and how those different elements interact with one another. If you need help with your website optimization, contact us and our strategy department will be more than happy to help!

More from the
DO Blog