How to A/B test a subject line in Klaviyo

To A/B test a subject line in Klaviyo, create a campaign, turn on the A/B test option, write two subject lines that differ in one clear way, choose a test share and a winning metric, then send. Pick the winning metric before you launch, and prefer clicks or revenue over opens.

Have this ready.

• A hypothesis, such as a benefit led line versus a curiosity led line
• An audience large enough to give a meaningful difference
• A winning metric chosen in advance
• A place to record the learning so it informs later sends

Steps.

1. 1. Write the hypothesis

   State what you expect and why, such as a specific benefit performing better than a vague teaser. A test without a hypothesis rarely teaches anything.

2. 2. Create the campaign

   Open Campaigns, create the email as normal, then switch on the A/B testing option for the subject line.

3. 3. Write two distinct variants

   Change only one thing between them, such as length, tone or including the product name. Keep the preheader and content identical.

4. 4. Choose the test share and winner rule

   Set what portion of the audience receives the test, how long to wait, and which metric decides the winner. Clicks or placed orders are more dependable than opens.

5. 5. Check the audience

   Make sure the segment is large enough and healthy. Very small groups produce noise rather than answers.

6. 6. Schedule and let it run

   Send or schedule, then leave the test alone until the wait time ends. Changing things midway invalidates the comparison.

7. 7. Record and reuse the learning

   Note the result and the lesson in a shared log. Apply it to later subject lines and retest occasionally, as audiences change.

Common pitfalls.

• Testing several changes at once, so you cannot say what caused the difference.
• Choosing a winner on opens when privacy features inflate that number.
• Declaring a winner from a very small audience.
• Never writing results down, so the team repeats the same test.

Check it worked.

• One variable differs between versions
• Winning metric was chosen before sending
• Audience size is reasonable for the test
• Content and send time are identical for both
• Result and takeaway are recorded

Klaviyo changes its interface from time to time, so treat the steps as the order of work and match the names to what you see on screen.

Rather have this done for you? See our klaviyo email marketing service or book a free audit.