Split Testing (or A/B Testing)
Split testing (A/B testing) sends email variants to separate groups and compares a chosen outcome, such as clicks or completed signups.
Split testing (A/B testing) sends two or more versions of an email to different parts of an audience and compares them on one chosen metric, such as open rate, click-through rate, or conversions.
What you can test
Subject lines, call-to-action buttons, images, copy length, and send time are the usual candidates. Change one element per test. If the versions differ in both subject line and layout, a win tells you nothing about which change caused it.
How to set up a test
Pick the metric before you write the variants. Use a sample-size calculator to find the minimum audience each version needs to show a real difference. A two-version test usually splits the test group evenly. Adjust the split when there are more variants. Many teams test on a portion of the list and send the winner to the rest, for example testing on 20% and sending the winning version to the remaining 80%. Account for list growth and churn during the test window, since the audience at the end may not match the audience at the start.
Common mistakes
Testing several variables at once, using samples too small to separate a real difference from noise, and calling a winner before enough results are in. A subject line that wins on opens can still lose on clicks or conversions, so judge the test on the metric you chose at the start.
Related: Subject line, Open rate, Click-through rate (CTR), Conversion rate, Email deliverability.