Overview
A/B tests allow you to create multiple versions of an email template in a sequence step to determine which message resonates more with prospects. With more than one email template, you can refine messaging and communication over time to book more meetings and close more deals.
Did you know? The more A/B tests a sequence has, the better for your sender reputation and deliverability score. Varied messages tend to indicate authenticity instead of spam. For more deliverability tips, check out our email deliverability checklist.
Check out the following sections to use A/B tests in a sequence.
Use an A/B Test in a Sequence
Some Apollo plans include A/B tests. To access A/B tests, upgrade your plan. If you have questions about upgrading, reach out to the Apollo sales team.
First, create a sequence or use an existing sequence.
Then, add an A/B test:
- Launch Apollo and click Sequences. Select a sequence.
- Add an automatic or manual email step, then click Add A/B Test on the step you want to test.
You can only A/B test email steps. It isn't possible to A/B test other sequence steps or an entire sequence flow.
- Apollo creates a new email with no subject or body. Click the email.
- Edit the email template or use the AI assistant to generate content. When finished, click Save.
To effectively identify what is and isn't working at speed and scale, begin by A/B testing wildly different email templates. Then iterate on the templates of the best-performing messages by A/B testing more specific variables like individual subject lines or body text. Learn more in lesson 5 of our master class on building a world-class outbound program.
- After editing your message, toggle the emails on.
- Apollo evenly distributes each message variant in the A/B test to different contacts in the sequence step.
You can repeat this process for additional emails and sequence steps.
If you use A/B tests in multiple sequence steps, Apollo randomly distributes message variants in each sequence step. If a contact receives version A in step 1, it doesn't necessarily mean they'll receive version A in step 2. You can't currently assign specific message variants to specific contacts in each step.
Over time, observe the delivery data Apollo collects for each email. When you've collected enough data, you can determine the best-performing messages and deactivate underperforming content.
Rather than A/B test an arbitrary number like 100 or 200 delivered messages, optimize by testing an amount that is statistically significant. Online calculators can help determine statistical significance. Then, only determine an A/B test winner if you have a 90% or greater confidence interval. Learn more in lesson 5 of the Build a World Class Outbound Program Master Class.
In general, higher performing content will have higher delivery and interested rates, and lower spam blocked and opt out rates.
