# Shopify Rollouts: how native A/B testing works (2026)

Source: https://trymimetic.com/guides/shopify-rollouts

[Guides](/guides) / Revenue QA

# Shopify Rollouts: how native A/B testing works

Rollouts is Shopify's own way to schedule a change or test it against what you have now, built into the admin. It arrived with the Winter '26 Edition for themes and has grown since: checkout and customer accounts in Spring '26, then catalogs and discounts in Fall '26. You'll find it under **Markets > Rollouts**.

Here is what it can do, what it can't, and how to get a clean result from your first test.

## The three kinds of rollout

| Type | What it does |
| --- | --- |
| **Launch** | Makes a change permanent at a set time |
| **Event** | Makes a change for a set period, then puts things back |
| **Experiment** | Shows two versions to two groups of visitors and compares them |

Launches and events are scheduling tools. Experiments are the A/B test.

## What you can change

- **Theme.** Edit the live theme, or swap in a different theme. One theme change per rollout.
- **Checkout and customer accounts.** Swap in a different configuration. One per rollout.
- **Catalogs.** Set market catalogs to active or draft, which is how prices and product availability change in a rollout. B2B catalogs can't be included.
- **Discounts.** Turn up to 250 existing discounts on or off.

You can't edit Liquid templates directly inside a rollout. Make the change on a theme copy first, then use that copy as the new version.

## Which plan you need

Rollouts run on the Basic plan and up. **Experiments, the A/B tests, need the Grow plan or higher.** On Basic you can schedule launches and events, but not split visitors into two groups.

## How visitors are split

Shopify puts each visitor in a group before the page loads, so nobody sees the page flicker from one version to the other. The group is stored in an essential cookie, so a returning visitor sees the same version until the cookie expires or is cleared. A new phone or browser can land in a different group.

A few details that matter when you read the results:

- **Groups are random.** You can't aim a test at a customer segment, a tag or a device type.
- **Two versions only:** the current one (control) and the new one (treatment).
- **Changing the split mid-test** only affects new visitors.
- **Visitors who decline tracking still see a version** but aren't counted in the results.
- **Bots are left out.** So is anyone who ended up in both groups in the same browser.

## What the results show

| Rollout | Results Shopify reports |
| --- | --- |
| Theme experiment | Conversion rate, bounce rate, add-to-cart rate, reached-checkout rate |
| Checkout experiment | Checkout conversion rate |
| Launch | Sales, orders, average order value, sessions, conversion rate and more |

Experiments include statistical results, and results update a few minutes after each visit.

One gap worth knowing: **a theme experiment reports rates, not revenue.** Conversion rate can go up while order value goes down, so before you call a test, check sales and average order value for the same period in your own reports.

## Limits to know before you start

- **Vintage themes can't be used.** Rollouts needs an Online Store 2.0 theme.
- **Online store only.** Headless storefronts, and checkouts outside the online store, can't be tested.
- **Your other tools don't know about the test.** Rollout and version IDs aren't passed to app pixels, custom pixels or Google Analytics. GA4 sees one blended site, and ads and product feeds see one price. A catalog test that shows some visitors a different price than your Google or Meta feed can get products disapproved, so keep price tests off feeds, or make them short.
- **You can't publish a theme while a theme rollout is unfinished.** That includes publishing from the Shopify CLI or another app.
- **Discount codes travel.** A code in an experiment can still be shared with people outside the test group.

## Rollouts or an A/B testing app?

Rollouts covers the most common store test, "is this version of my theme better than the current one?", with the split handled by Shopify before the page loads. It has no extra script, and your results sit next to the rest of your Shopify analytics.

A testing app still makes sense when you need what Rollouts doesn't do: more than two versions, tests aimed at a segment or device, or tests on parts of the page you'd rather not make a whole theme copy for.

Whichever you use, watch out for tests that send one group through an extra redirect to reach their version. That group waits for an extra page load the other doesn't, and the result gets tilted against the change.

## What to test first

Rollouts is only as useful as the change you put in it. A good first test fixes something shoppers are actually tripping over: a product page that hides the price or shipping cost on phones, an add-to-cart button that gives no sign it worked, a cart that surprises people with costs at checkout. Find those first, then test the fix. For what other stores test most, see [what to A/B test on an online store](/guides/what-to-ab-test).

## How Mimetic works with Rollouts

Mimetic checks your store for the problems that cost sales, then builds the fix on an unpublished copy of your theme. Your live theme isn't touched. From there you choose: publish the copy, or test it with **Shopify Rollouts**. Mimetic gives you the copy's name and opens Rollouts in your admin, where you start the experiment and read the result.

> [Get Mimetic on the Shopify App Store](https://apps.shopify.com/mimetic) to find the first change worth testing.

## Common questions

**Is Shopify Rollouts an A/B test?**
An experiment rollout is. Launches and events only schedule a change.

**What plan do I need for Rollouts A/B tests?**
Grow or higher. Basic can schedule launches and events.

**Can Rollouts test prices?**
Through catalogs, yes: you can activate a market catalog with different prices for part of your traffic. Keep an eye on ads and product feeds, which only show one price.

**Does Rollouts slow down my store?**
Shopify puts each visitor in a group before the page loads, so the page doesn't flicker between versions.

**Can I see Rollouts results in Google Analytics?**
Not directly. Shopify doesn't send rollout or version IDs to GA4 or to pixels, so read results in Shopify.

### See what is costing your store sales.

[Scan my site](/)
