Email A/B tests

A/B testing in email is easy: split list, send variant A to half, variant B to other half, measure.

What to test

• Subject lines (highest leverage)
• From name
• Send time
• Email copy (length, tone)
• CTA

Sample size

Most email tools build in sample-size calculators. Rule of thumb: test needs 500+ recipients per variant for meaningful data.

Single variable

One test, one variable. Don't change multiple things at once.

What to do with this

• Test one variable at a time, multi-variable tests contaminate the learning and produce false attribution
• Test offer and subject line first, they produce the biggest revenue swings, button-color optimization is a distraction
• Require sample size before calling a winner, email A/B tests on 500 subscribers produce noise not signal
• Run tests for full send cycle, not just overnight, opens and clicks lag 48-72 hours on most B2B lists
• Document winners in a rolling doc, same bad subject line re-tested in 6 months is a preventable time tax