A/B/n Testing

What is A/B/n testing (multivariate testing)?

A/B/n testing is best described as an extension of typical A/B testing. 

In regular A/B testing (or split testing), you compare the performance of two different variants of a webpage by showing one version to one group of customers and the other to a different group. The goal is to see which version has higher conversion rates.

In A/B/n testing, you can compare the performance of more than two versions of a website. The little n refers to the number of variants you want to test: you could test 3 variants (A/B/C test), four variants (A/B/C/D test), and so on.

Je7wtGx9j1SGENEsAfRANo7ThC u5iSRJUCASrU2mkNJipVRda98YPgkvZG7tOjowgqqoITuPhENRYGT7H16BmT1KzvZUP8 4bm OEymC0tiCFmB7DTbEcZLoVyJdAIjLXAHENwF01iLuFsmdJkiOak T1KF2C46JtTfY03tPxAyeL8AmPaq4VGZFWD 0A - A/B/n Testing

Why would you want to test more than two versions?

A/B/n testing helps marketers gain a much better understanding of what works and what  doesn’t when it comes to a landing page or website. You don’t have to limit yourself to testing just two versions of a page at once, you can test additional versions simultaneously.

For example, if you’re creating a landing page for an upcoming event at your company, you can test multiple value propositions to see which one converts best.

With the A/B/n test results, you can make significant changes to your landing page based on data rather than gut feelings.

A/B/n testing example

Check out how Obvi A/B tested different unique selling propositions (losing weight vs resting better):

5iaE4aIJzbLZ2453nP 5QqwtKZb0jST43WWbBqVxRMoXgvhqvVCJOEzK0wHpJ3vzeeyYYGXB22Y2dD66exI0idIVdFJB5evJk5XBNeHOTbcYMGBH0eKIhMD Y8WZ2V4yIDxe17qFDLhrhXR7TS  BkxoIDW03IbEACxLc9b3TO4f6joAKNrC1eEWNd8KIQ - A/B/n Testing
lNw2cFF 7owl82T2ySuKYa oGrgZYgEfdAaHt81 7biXEp3mQID82XMDunHOt0hBznmfChIpyIr5g3Zk egDC5nEBFtkIEonWRTp8cdl uIMhakrKuCD0Zlkr0gvyJmyYmXe RR5ooxpwg7 MmpHCvR3oSYZ hqkDlamMizzUwIh5UlOBUZomK1Ts74YZw - A/B/n Testing

A/B/n testing downsides

There are a couple things to watch out for when it comes to running an A/B/n test. First of all, make sure to keep your variants at a reasonable level. If you have too many, you’re probably doing a whole bunch of work for nothing.

Another potential problem is that the more variants you’re testing, the more traffic you’ll need to reach statistical significance. This increases the amount of time you’ll need to wait to see which version actually has the best conversion rate.