A/B Testing

Version 5.0 offers an additional feature using AI to compare the potential effectiveness of two different subject lines.  Known as Ai/B testing, more information can be found in this article.

Any experiment with one or more test groups and a control group is in effect an A/B test; the purpose is to observe and measure the difference in the effect. When measuring your internal communications, you may want to experiment with different variables (such as sending from a different From address or at a different type of day) to see which will produce the highest results. Here is a summary to help you successfully conduct A/B testing on your email messaging:

Start with your hypothesis

A hypothesis is “an assumption about certain characteristics of a population.” The hypothesis defines your test and outcome.

As a communicator, you most likely know your audience and what is important to them (If not, consider conducting a simple survey or do some analytics analysis to find out.)

With that in mind, develop a testable hypothesis, such as “We expect that sending from the leader’s name, instead of a shared mailbox, will cause our attention rate to go up”. To conduct the test, first you’ll need to decide what element you are testing, and what metric will serve as your measure. You might use the open rate or attention rate when you are testing From addresses or subject lines, or the click through rate when you are testing different calls to action.

Create your sample sizes

Next, you will need samples of your list (A and B), which should be randomly selected and not overlap. You can use PoliteMail's list operations to narrow down your list populations.

Measure your results

Now use PoliteMail's reporting tools to compare your metrics, and see if your hypothesis proves correct or not. PoliteMail provides tools such as Comparison and Segmentation to help compare your results.

For more information about A/B testing and how to generate accurate, randomized samples, please visit our blog.