The built-in A/B test in Campaigns lets you compare subject lines, sender names, and email content – but it sends both versions at the same time. This means you can't use it to test different send times or compare entirely different content in a granular way.
This guide shows how to split contacts randomly using an Automation workflow, so you can send campaigns at different times or with different content variations and compare results.
The process has two stages:
Before You Begin
This method is available on all Omnisend plans.
You need at least 10 contacts subscribed to email for reliable results. 2,000+ contacts are recommended for statistically meaningful data.
You need access to Automation workflows to complete this setup.
The A/B test block in Automations creates a 2-way split only. See the FAQ below if you need 3 groups.
Stage 1: Creating a Segment
The Automation randomly assigns contacts to two groups using tags: variant_a and variant_b.
Step 1: Create a segment
Create the segment you want to use as the Automation trigger first.
Go to Audience → Segments → Create segment → Create from scratch.
Add at least one filter: Subscribed to Email channel. If you want to target a specific list, add additional filters or import contacts directly to the segment.
Save the segment.
💡 If you're new to segments, see Create & Use Segments in Omnisend.
Step 2: Set up the Automation Workflow
Go to Automation → Create workflow → Create from scratch.
Under the trigger settings, select Contact enters a segment and choose the segment you created in Step 1. Enable the option to include contacts already in the segment.
Add an A/B Test Split block to the workflow:
On Path A, add an action to assign the tag
variant_a.On Path B, add an action to assign the tag
variant_b.
Step 3: Trigger the workflow
Once your automation is ready, click Start workflow.
Make sure you have checked "Trigger workflow for contacts in the segment" in Step 2. This way, contacts currently in the segment will enter the workflow when you turn it on.
💡 Don't see the checkbox?
The "Trigger workflow for contacts in the segment" option only appears for workflows that have never been enabled before. If you already turned the workflow on at any point, duplicate it and enable the checkbox in the new draft before activating it. Alternatively, you can re-trigger the workflow manually. See Contact Enters/Exits a Segment – Re-triggering the Automation.
Once the workflow runs, each contact is tagged with variant_a or variant_b. Disable the workflow after all contacts have been processed.
Stage 2: Sending Campaigns to Each Segment
Step 1: Create Segments Based on Tags
Create two new segments:
Segment A: filter by tag
variant_aSegment B: filter by tag
variant_b
Step 2: Send the Campaigns
Create two copies of your Email Campaign – each with the content or send time you want to test. Send or schedule:
Version A → Segment A (
variant_a)Version B → Segment B (
variant_b)
📌 Planning to run this test again in the future? Don't reuse the same tagged segments without refreshing them first. Your audience changes over time, and reusing old tags means contacts aren't randomly reassigned.
Before each new send:
Go to Audience → Segments and open the segment for
variant_atag.Select all and use bulk actions to remove the tag.
Repeat for the
variant_btag.Go back to your workflow and re-trigger it so contacts are reassigned freshly (see Step 3).
Step 3: Review results
After both campaigns are sent, go to Reports → Campaigns → Campaign performance and select each campaign to compare open rates, click rates, sales, and other metrics.
FAQ
What's the difference between this method and the native A/B test in Campaigns?
The native A/B test automatically splits your audience, sends both versions simultaneously, and selects a winner based on open or click rate. Use it when you want to test subject lines or sender names quickly. Use this segment method when you need to test different send times, compare more complex content variations, or have more control over how results are evaluated.
Can I use this method to test different send times?
Yes – this is the recommended approach. The native A/B test sends both versions at the same time, so it cannot compare delivery timing. Split your contacts using this method, schedule separate campaigns to each tagged segment at different times, and compare results in reporting.
Can I split contacts into 3 groups?
Yes. Add a second A/B Test Split block on one of the two paths (for example, Path B). For an approximately equal 3-way split, set the first split to 33/67, then set the split on Path B to 50/50. This creates three branches of roughly equal size. Assign a different tag to each path (variant_a, variant_b, variant_c), create a segment per tag, and send a separate campaign to each group.
If you have any questions or need assistance, do not hesitate to contact our support team at [email protected] or via in-app chat.