Read more posts by this author.
A Quick Guide To Deciding The Right Percentage And Time For A/B Testing.
SendX allows you to do A/B Testing for multiple variations of Subject Lines and From Names to figure out which one performs better. As a marketer, we always need to keep on testing and analysing the results, and then adjust accordingly. This A/B Testing feature in SendX allows you to take out a certain percentage of your contact base (at random), and test out multiple variations for a certain duration. After that, SendX will automatically evaluate the results and send out the winner.
- Create a new A/B Test (Email Subject or From Name).
- Under the Content tab, you will see two fields labelled Percentage of contacts for A/B testing and Wait before deciding for. These are the fields that are used to decide the right percentage to segment for A/B testing, and the time to set the multiple variations before evaluating the result.
Percentage of contacts for A/B testing
This drop-down menu allows you to set the percentage of your subscriber base that SendX will segment out randomly, and then test the variations on.
For ex- If you select 20, SendX will take 20% of your subscriber base at random, and test out the variations you would have set. And the variation which performs the best on these 20% is what is sent out to the rest of the 80% after the test ends.
Wait before deciding for
The two fields under this, allow you to set the duration for which the A/B Test would run. So whatever numeric value you and one option from day/hours/minutes you choose, SendX will run the A/B test on the chosen percentage of contacts for that duration. After that, it will analyze the results and then send out the email accordingly. For ex - If you select 3 hours, SendX will take your chosen percentage of the subscriber base, and do the A/B test. Depending on the results it receives within 3 hours, it will choose the best performer and send it out the to remaining contact base.
And that's it! Using these two fields, you can decide the right percentage and time for the A/B test to run in SendX.