Shopify Rollouts Lets You A/B Test Your Theme, With Limits

Shopify now lets you schedule a theme change, run a sale that switches itself off, or split traffic between two versions of your store, all from the admin and without a testing app. The feature is called Rollouts, and on 22 September 2026 Shopify redesigned how you set one up. It is more widely available than most merchants assume, and it has limits worth knowing before you trust a result. This guide covers both.

Everything below is from Shopify's changelog and Help Centre, checked on 28 September 2026.

What is Shopify Rollouts?

A scheduled set of changes to your store that you can release to everyone, to some visitors, or for a fixed period. In Shopify's words from the Rollouts help page, "a rollout is a scheduled set of changes to your online store's main theme, your checkout and accounts pages, your product catalogs, or a combination of these." You find it under Markets, then Rollouts, in the admin.

The 22 September changelog entry added a redesigned setup flow, the ability to "make changes to your theme, checkout and customer account pages in a single rollout," a view of what share of traffic a rollout reaches and how it is split between test versions, and conflict detection when two rollouts touch the same thing.

What are the three types of rollout?

Shopify's rollout types page defines three, and choosing the right one is most of the work.

• Launch. "The changes are merged into your store permanently." Use it to schedule a new theme or a redesigned page to go live at a set time, instead of someone clicking publish at midnight.
• Event. Publishes changes "for a set period, and then rolls them back when the rollout ends." This is the one most stores will get the most from.
• Experiment. Compares "control and treatment groups to evaluate the effectiveness of changes." This is Shopify's native A/B test.

A practical note on launches: schedule them for a quiet hour on a weekday rather than the moment a campaign starts, and have someone check the live store straight after. The value of a scheduled launch is that nobody has to be awake at midnight to press publish, not that nobody needs to look. If something breaks, you want the person who can fix it to be at their desk when it goes live.

Which Shopify plans can use Rollouts?

More than most people think. Shopify's requirements page states that "Rollouts are available on the Basic plan or higher" and "Experiments are available to stores on the Grow plan or higher."

So scheduled launches and self-reverting events are open to nearly every store, and theme A/B testing does not need Shopify Plus. For many Indian D2C brands on Grow or Advanced, that removes the monthly cost of a third-party testing app for straightforward theme tests.

There are technical conditions. The same page says "You can't apply rollouts to vintage themes" and "You can't change Liquid templates as part of a rollout." If your store still runs a pre Online Store 2.0 theme, Rollouts is not available to you until the theme is replaced. Headless and custom storefront checkouts are not supported for theme and checkout changes either.

How should Indian stores use events for sales?

For the thing that goes wrong every festive season: the sale theme that nobody switches back.

A Diwali or end of season sale usually means a new banner, a changed homepage, sale badges and sometimes a different announcement bar. The usual method is to duplicate the theme, edit it, publish it at the start of the sale and remember to publish the old one afterwards. That last step is where it breaks, at the end of a long week, often late at night.

An event does the switching both ways. You build the changes, set the start and end, and Shopify rolls them back when the event ends. Two cautions from Shopify's documentation apply:

• Catalog changes behave differently from theme changes. When a catalog rollout reaches 100 percent of visitors, it updates prices and availability across the online store, checkout, cart, headless storefronts, Shopify POS and Google and YouTube feeds. And "any edits that you make to a catalog while the rollout is running stay in place" after a pause or rollback. A price you fix by hand during the sale does not revert on its own.
• Test the event on a preview before the sale, not during it. Theme changes in a rollout cannot touch Liquid templates, so anything that needs code must already be in the theme before the event is built. If a sale feature needs development work, that has to happen weeks ahead, which we cover in our guide to updating a customised theme safely.

How do Rollouts experiments work?

An experiment sends part of your traffic to the changed version and the rest to your current store, then reports how each group behaved. According to Shopify, "by default, a new experiment uses a traffic split of 50% control and 50% treatment, and an end date 90 days after the experiment starts." Both can be adjusted.

Experiments, and any rollout that reaches less than 100 percent of visitors, apply only to the online store. Shopify's rollout analytics page lists what gets measured:

• Theme experiments: conversion rate over time, bounce rate, reached checkout rate and add to cart rate.
• Checkout experiments: checkout conversion rate.
• Combined experiments: the storefront rates plus conversion rate over time.

What do Rollouts experiments not tell you?

They do not report revenue. Average order value and gross sales are not part of experiment reports, even though they are available for launches. Every metric an experiment shows you is a rate.

That matters more than it sounds. The easiest way to lift conversion rate is to make buying cheaper or smaller: a bigger discount, a free shipping threshold that drops, a cheaper product pushed higher on the page. All three can win an experiment on conversion rate and lose money at the same time. A variant that converts better but lowers average order value is not a winner until someone has checked the orders behind it.

Three habits keep results honest:

1. Decide the success measure before you start, and if it involves revenue, plan to pull order data for the experiment period separately.
2. Change one thing per experiment. Shopify lets you combine theme, checkout and account changes in one rollout, which is useful for launches and risky for tests. When three things change together, you cannot say which one moved the number.
3. Do not stop at the first good week. Shopify's analytics page does not state a minimum run time or sample size, so the discipline is yours. Run through at least a full weekly cycle, and longer for lower traffic stores, before reading anything into a difference.

Does the session measurement change affect experiments?

Yes, for anything that ran across it. Experiment metrics are session based, and Shopify changed how sessions are counted between 21 and 23 September 2026. Both groups were measured the same way on any given day, so a test comparison is less exposed than a before and after chart. But an experiment that started before 21 September has results measured under two methods, and its conversion rate over time chart will show a step that has nothing to do with your change. We explain what moved in our breakdown of the session measurement update. If in doubt, restart the test.

When is a testing app or a developer still the better choice?

When you need to test something Rollouts cannot touch, or measure something it does not report.

• Template or code changes. A new product page layout that needs Liquid work cannot be built inside a rollout. It has to be built into the theme first, then tested.
• Revenue based decisions. Price tests, bundle tests and anything judged on average order value need order level analysis that experiment reports do not provide.
• Headless storefronts. Rollout analytics measure the online store, cart and checkout only, so a Hydrogen or custom frontend needs its own testing setup.
• Stores on Basic. Launches and events work, but experiments need Grow or higher.

For most stores the right move is to use Rollouts for scheduling and simple theme tests, and to spend testing effort where the money is. Which parts of the store deserve a test first is usually clear from an audit, and what we find in a Shopify CRO audit shows where the biggest leaks tend to be.

Want help planning your first rollout?

Free test plan. Get in touch or email hello@exactwhy.com with the subject "Rollouts" and your store URL. Tell us your plan and what you want to test or schedule. We reply within 4 hours with the first experiment worth running, how to measure it properly, and whether your theme is ready for Rollouts.

Testing programme. If you want a steady run of tests rather than a one-off, we build the variants, run them through Rollouts or a dedicated tool where needed, and read the results against real orders, not just rates.