Campaigns

Group related experiments so you can plan them as one initiative and read their results side by side.

What a campaign is

A campaign is a container that groups related experiments. Each experiment still runs on its own — its own variants, traffic split, targeting, and schedule — but the campaign ties them together so you can see them as a single effort rather than a scattered list of tests.

Use a campaign whenever several experiments share a theme or a goal, for example:

A product-page redesign broken into a template test on the PDP, a JavaScript test on the add-to-cart button, and a URL test on a landing page.
A seasonal push where you test the same idea across collection, product, and cart pages.
An ongoing optimization program where you want one place to track everything you have shipped for a given surface.

Campaigns are organizational. Grouping experiments does not change how any individual experiment assigns visitors, serves variants, or computes its own results.

Every test type can belong to a campaign — URL redirect, template, theme, and JavaScript API tests can all be grouped together in one campaign.

Creating a campaign

Create the campaign first, then attach experiments to it.

Open the AB Test app in your Shopify admin and go to Campaigns.
Select Create campaign and give it a clear, descriptive name — this is how it appears in reporting, so name it for the initiative (for example, "PDP redesign Q3").
Save the campaign. It starts empty, ready for you to add experiments.

Adding experiments to a campaign

You can associate an experiment with a campaign when you create it, or add an existing experiment to a campaign afterward.

When creating an experiment — choose the campaign it belongs to as part of setup.
For an existing experiment — assign it to a campaign from the experiment's settings.

An experiment can be launched, paused, or apply its winner independently of the others in the campaign. Ending or applying one experiment does not affect the rest of the group.

Keep each experiment in a campaign focused on a single change. A campaign is the right place to hold several distinct experiments; it is not a way to bundle unrelated changes into one test, which would make results impossible to read.

Reporting across a campaign

The campaign view brings the results of its experiments together so you can review the whole initiative in one place. For each experiment you see its per-variant metrics — conversion rate, revenue, and the Bayesian probability to be best — without opening each experiment separately.

Results are still computed per experiment. AB Test does not merge visitors or conversions across experiments into a single combined statistic, because each experiment has its own variants and audience. The campaign gives you the summary and the context; the winner is still decided within each experiment.

From here you can:

Compare how each experiment in the initiative performed at a glance.
Open any experiment for its full breakdown, including segment results by device, new vs returning, country, UTM, referrer, and cookie.
Apply the winning variant of an experiment in one click when its results are conclusive.

For how the underlying statistics are calculated, and how revenue is credited to each variant, see Results and statistics and Revenue attribution.

Plan limits

The number of campaigns you can create depends on your plan.

Plan	Campaigns	Running tests
Base — $19/mo	1	2
Premium — $29/mo	5	5
Enterprise	Custom	Custom

On the Base plan you can maintain one campaign at a time; Premium allows up to five. If you reach your limit, remove or repurpose an existing campaign, or move up a plan. See Plans for the full comparison, including test types and tested-visitor allowances.

Next steps

Results and statistics — read Bayesian results and segment breakdowns for each experiment.
Audience targeting — control who each experiment in the campaign runs for.
What to test — ideas worth grouping into a campaign.
Plans — campaign, running-test, and visitor limits per plan.