Use this article to limit a Rebuy A/B test to one group of shoppers with Audience Targeting, so only shoppers who match your audience are split between your control and your variations. You can target shoppers by device, by new or returning visitor, by country, by Shopify customer tag, or by Klaviyo segment. Shoppers outside the audience see your store's default experience and are never counted in the experiment's results.
What Audience Targeting does in an A/B test
Audience Targeting controls which shoppers are eligible to enter a Rebuy A/B test. When a shopper reaches a page where the experiment runs, Rebuy checks them against the audience. A shopper who matches is assigned to the control or a variation using the traffic percentages you set, and a shopper who does not match is left out of the experiment entirely, with no variation assigned and no visit recorded.
The Audience section appears in the experiment editor for Widget, General, Checkout Offer, Smart Cart, and Global Smart Flow experiments. It defaults to Everyone (All shoppers will be entered into the experiment), which runs the test on all of your qualifying traffic. Experiments created before Audience Targeting was released on September 30th, 2026 keep the Everyone setting and run exactly as they did before.
Because shoppers outside the audience are never entered, the Visitors column on the A/B Testing dashboard counts only shoppers who matched the audience.
Add an audience to an A/B test
You add an audience to a Rebuy A/B test in the experiment editor, in the Audience section between Select Control and Variations and Experiment Goal. To add one:
From the Rebuy admin, go to Flows & A/B Testing > A/B Testing.
Click Create New Experiment to build a new test, or open an existing experiment that is still a Draft.
Scroll to the Audience section, below your control and variations.
Open the audience dropdown and select one option: Device, Visitor type, Location, Customer tag, or Klaviyo segment.
Set the value for that option in the field beside the dropdown. Each option's value is explained in its own section of this article.
Finish your experiment goal and test duration, then click Save Experiment.
Start the experiment. The audience is locked from this point until the experiment ends.
If you selected Klaviyo segment, Rebuy checks the segment when you start the experiment and blocks the start with an error if the segment is missing, not synced, or empty.
The following table compares the Audience Targeting options, with a column for what you choose beside the dropdown, a column for when Rebuy can first enter a matching shopper into the test, and a column for how shoppers who are not logged in are handled.
Audience option | What you choose | When a matching shopper can enter the test | Shoppers who are not logged in |
Device | Desktop, Mobile, or Tablet | First page view | Not affected |
Visitor type | New visitor or Returning visitor | First page view | Not affected |
Location | A country | Second page view onward | Not affected |
Customer tag | A Shopify customer tag | First page view | Left out of a tag audience, since guests have no customer tags |
Klaviyo segment | A synced Klaviyo segment | Second page view onward | Left out unless Rebuy can identify the shopper |
Target shoppers by device
The Device audience option limits a Rebuy A/B test to shoppers browsing on a Desktop, Mobile, or Tablet device. Select Device in the audience dropdown, then click the device type you want to test in the buttons beside it. Rebuy checks the device on the shopper's first page view, so a Device audience starts counting shoppers right away.
Target new or returning visitors
The Visitor type audience option limits a Rebuy A/B test to new visitors or to returning visitors. Select Visitor type in the audience dropdown, then click New visitor or Returning visitor beside it.
Visitor type is decided by the shopper's browsing sessions on your store, not by their order history, and a browsing session ends after 30 minutes without activity. A shopper who has ordered from your store before can still be classified as a New visitor. A shopper who enters a New visitor test stays in that test for up to 14 days, even after they come back as a returning visitor.
Target shoppers by country
The Location audience option limits a Rebuy A/B test to shoppers in the country you choose. Select Location in the audience dropdown, then search for and select a country in the Search for a country... field beside it.
A Location audience finds the shopper's country from their IP address, not from the country set for the shopper's Shopify checkout. A shopper browsing through a VPN is located by the VPN's IP address.
Note: A Location audience does not count a shopper's first page view. The shopper's location has to be looked up over the network, which cannot finish before the page renders, so Rebuy measures the shopper from their second page view onward. A shopper who leaves after one page is never entered into the test, so a Location audience reaches slightly fewer shoppers than a Device or Visitor type audience.
Target shoppers by Shopify customer tag
The Customer tag audience option limits a Rebuy A/B test to shoppers whose Shopify customer account carries the tag you enter, such as VIP or wholesale. Select Customer tag in the audience dropdown, then type the tag in the Type a customer tag... field beside it.
Customer tags belong to a logged-in customer's Shopify account, so Rebuy can read them only once the shopper is logged in. A guest shopper who is not logged in has no customer tags and is left out of any Customer tag audience that requires a tag. A guest who logs in after being left out keeps that decision for up to 14 days.
Target a Klaviyo segment
The Klaviyo segment audience option limits a Rebuy A/B test to shoppers who belong to a segment in your connected Klaviyo account. Select Klaviyo segment in the audience dropdown, then search for and select the segment in the Search Klaviyo segments... field beside it. The list shows the segments synced from Klaviyo to Rebuy. To set up that sync, see Add Klaviyo Segments to Rebuy Data Sources.
Rebuy checks a shopper against a Klaviyo segment only when it can identify who the shopper is, such as a logged-in customer or a shopper Klaviyo already recognizes. A guest shopper Rebuy cannot identify is treated as not in the segment, is left out of the test, and sees your store's default experience.
When you start a Klaviyo segment experiment, Rebuy confirms the segment exists and is synced with at least one member. If it is not, the experiment does not start and an error explains why. Fix the segment sync, then start the experiment again.
Note: A Klaviyo segment audience does not count a shopper's first page view. The segment has to be looked up over the network, which cannot finish before the page renders, so Rebuy measures the shopper from their second page view onward. A shopper who leaves after one page is never entered into the test, so a Klaviyo segment audience reaches slightly fewer shoppers than a Device or Visitor type audience.
How long a shopper keeps their audience decision
Rebuy makes the audience decision for a shopper once, the first time it can check them on a page where the A/B test runs, and keeps that decision for 14 days, the lifetime of the experiment cookie. A shopper who matched the audience keeps their assigned control or variation for those 14 days, and a shopper who did not match stays out of the test for those 14 days.
Rebuy does not re-check the audience when a shopper's details change during those 14 days. A guest who logs in, a new visitor who comes back, or a customer who gets a new tag keeps their original audience decision and their original variation.
Change the audience on a live A/B test
The audience on a Rebuy A/B test can be edited only while the experiment is a Draft. Once the experiment is live, the Audience section is locked until the experiment ends, which keeps the group of eligible shoppers the same from start to finish so the results stay comparable. To test a different audience, end the live experiment and create a new experiment with the new audience.
When a variation has its own targeting
Audience Targeting decides who enters a Rebuy A/B test, and any targeting built into a variation still runs on top of it for the shoppers who get in. A shopper has to pass the audience and the variation's own targeting to see that content.
A Global Smart Flow variation that contains its own targeting components, such as Klaviyo Segments, Customer Tags, New Users, Returning Users, or Location, filters shoppers again inside the variation after the audience admits them. For example, with a United States Location audience and a flow variation whose Location component targets California, only shoppers in California see that variation's content. Rebuy shows a warning when you start a Global Smart Flow experiment set up this way, and you can still start it when the layering is intended.
A widget variation whose data source uses a customer tag, Klaviyo segment, or location data source rule applies that rule after the audience, so the rule can further narrow which products an admitted shopper sees.
A checkout Progress Bar limited by country uses the country set for the shopper's Shopify checkout, while a Location audience uses the shopper's IP address. A shopper can pass a Location audience and still not see a country-limited checkout Progress Bar when those countries differ.
What Audience Targeting does not do
Audience Targeting accepts one audience option per experiment. You cannot combine options in one audience, such as mobile shoppers in the United States, and you cannot use OR or NOT logic between options.
Audience Targeting decides who enters an A/B test and does not show different variations to different audiences within one experiment. To compare how different audiences respond, run a separate experiment for each audience. A Smart Cart or Global Smart Flow experiment cannot run alongside another experiment of the same type, so for those types run the audience experiments one after another.
Audience Targeting does not preview how many shoppers an audience will reach before you start the test, and it does not target by shopping behavior such as pages viewed, time on site, or cart subtotal.
Plan a longer test for a smaller audience
A Rebuy A/B test limited to an audience collects visitors more slowly than the same test run on everyone, because only matching shoppers are counted. A test that reaches 20% of your traffic takes longer to reach a confident result than one that reaches all of it. For a store with lower traffic, start with a Device or Visitor type audience, since Device and Visitor type audiences are checked on a shopper's first page view and usually cover a large share of traffic.