People argue about subject lines for hours. A test settles it in a week: two variants, recipients split in half, the numbers pick the winner. This guide is about what to test, for how long and how to read the result.
Test one thing
Question vs statement. Company name in the subject vs without. Short vs longer. Not three changes at once, or you will not know which one worked.
Enter variant B in the campaign
In the sequence, at the first email, click "Subject line A/B test" and enter the second variant. Recipients are split in half automatically; everything else (copy, timing, mailboxes) stays the same.
Wait for the sample
At least 100 companies contacted per variant. A 3% difference on 30 companies is one opened message, not a result. At 100 companies a day you have a result in two days.
Read opens and replies
A subject line with more opens but fewer replies is lying: it triggered curiosity but did not promise what the email delivers. The winner is the variant with more replies; opens are just a hint.
Lock in the winner and test the next
The winner becomes variant A of the next campaign and you test a new idea against it. After five rounds you have subject lines proven to work in your market.
- One variable per test
- Variant B entered in the campaign
- At least 100 companies per variant
- Evaluated 48 hours after the last send
- Winner by replies
- Result written down and used in the next campaign