A/B testing

Version 4.6 of CiviCRM introduced a new tool that will let you create different versions of a mailing and compare how effective they were based on the response statistics. You can vary the differences based on the From address, subject line, or the entire e-mail.

When constructing an A/B test, it replaces the process previously described for creating a standard bulk e-mail. In the first step, you name the mailing and decide which of the three test types you will use. In the second step, you define your recipient list and target audience breakdown. The recipient list is selected in the same way as you do a normal mailing. The target breakdown determines the percentage of recipients that will randomly be selected to receive the A version of the mailing, the B version of the mailing, and the Final version of the mailing. The concept here is to send different versions of the mailing to a small subset of your recipient list to determine the version that generates a better response...