A/B testing is the experimenting of 2 or more campaigns to gain knowledge of which campaign should be sent to a larger group. Two or more versions are compared, which are identical except for one variation that might affect a user's behavior. JangoMail offers an easy-to-use A/B Testing feature. 

Setting up an A/B Test involves a few steps: 

1. Create the list you will use, if you do not already have one, or setup a filtered extraction.
2. Create the email messages that you will use. 
3. Setup the A/B Test Campaign.
4. Review the Reports.  

How to Set Up an A/B Test

Step 1:  Create list or filtered extraction to be used with A/B Test

1. Create and/or choose the list you will use in the A/B Test
2. Optionally, you can filter your list to send an A/B Test. If using this option, you must create your A/B Test campaigns before filtering your list

Step 2: Create the campaigns for your A/B Test

1. Navigate to the Campaigns section
2. Create a message, leaving the To Lists tab set to no lists selected
3. Type in a subject line, the content of the message, and click Save

Pro Tip: add a Tag to help identify your A/B Testing Messages. 

Step 3: Set up the A/B Test

1. Go to Campaigns > select New Message
2. Select A/B Test as your new message type
3. Name your A/B Test, select your list, select the messages you wish to use, and click Next
4. Select the percentage of recipients that should get each message and click Next
5. To send the A/B Test, click SEND THIS JOB NOW 

Step 4: Review reports

Once your A/B Test is finished sending, you can review the results in Reports. Before reviewing, allow time for recipients to receive the campaigns. 

1. Go to Reports > click OPEN CAMPAIGN DASHBOARD
2. Find your A/B Test "parent message" and expand to see results
3. Compare results of the A/B Test messages (aka the "child messages")

When viewing Reports, you will find your Recipient, Open, Click, Bounce, Unsubscribe, Complaint, Block data and SMTP Log details under the "child messages". The "parent message" will show the total counts, but will not show the raw data, with the exception of Recipients. Preprocessing results are under the "parent message".