Let AI help you do more of what works
Marketing teams who drive consistently strong performance frequently test and iterate on their content strategy through A/B testing. Klaviyo AI enables you to run smarter A/B tests by calculating the statistical significance of your test results, giving you win percentage estimates, and automatically directing recipients to the best-performing content. See how it works across various types of A/B tests in Klaviyo.
Campaigns
When you run an A/B test on message content in campaigns, you can choose how much of your audience should participate in the test. By doing this, you can show the winning content variation to the most subscribers. To determine a winner, Klaviyo uses machine learning to observe the patterns of how recipients interact with the test variations.
Flow messages
When you A/B test message content inside of a flow, Klaviyo will use AI to automatically distribute traffic toward the winning variation as it gains confidence in the result. This allows you to serve the best version of content to the most recipients.
Sign-up forms
When you A/B test a form, you can allow Klaviyo AI to select the winner once statistical significance is reached, and then automatically deliver the winning content variation to the remainder of your audience. This way, you can continuously optimize your content without sacrificing conversions by letting AI show the best performing variation to the largest possible audience.
Always be testing (with smart A/B tests)
Explore the different types of smart A/B tests that you can run in Klaviyo. While you should test often, be sure that you are never testing more than 1 variable at a time.
Email subject lines
You can test email subject lines for both email campaigns and emails that are inside of flows. Test subject line to see which copy generates the highest open rate.
When you run an A/B test on subject lines, you should keep the content inside the messages constant between variations. Changing the content in addition to could cause a difference in deliverability between variations, which may skew your results or make them less reliable.
Campaign content
You can test email and SMS campaign content to see what your subscribers want to click on.
Note that if you run an A/B test for email content, you should not change the subject lines between the variations. Hold the subject line constant so you can measure the impact of the content.
Flow message content
Just like message content in campaigns, you can also test email and SMS content in your flow messages. This will help you understand which content in your flow messages is generating the most clicks or most attributed orders.
Sign-up forms
You can test form design, incentives and offers, or display options using an A/B test. Test different variables to see if you can increase your form submit rate.
Campaign send time
For both email and SMS campaigns, you can test different send times to see if they show a lift in engagement around a particular time of day.