AN A/B TEST FOR CHECKOUT ABANDONMENT

Summary from the TLDR Marketing newsletter

This post breaks down an example of a checkout abandonment A/B test that tested a delay of 1 hour vs. 2-3 hours when sending the first follow-up email. Brands will often focus on one metric, which is the number of placed orders for the first email. The 1 hour test version will always win — but the truth is this doesn’t paint the full picture of what’s happening. A proper test would split the following into 2 cohorts, tag 50% of the profiles as 'control' and the other half as 'test', and capture the flow start date. Segments to pull include the number of people in each cohort in the last month, discounts used by cohort, placed orders by cohort, and unsubscribed by cohort.