Illustration of a ruby gem on a timeline that forks into two lanes, one with a storefront and one with a price tag, representing Shopify Rollouts scheduling and A/B testing

By Lance Allison, Founder of Ruby Digital Agency | October 2026 | ~24 min read

This morning Shopify sent me an email with a dry subject line about rollout schedules and traffic allocation on discounts. It was addressed to app developers, and most merchants will never see it. I think that is a shame. Buried in a notice about GraphQL scopes is one of the more useful things Shopify has given store owners this year: you can now put your discounts in the same scheduled, reversible, testable package as your theme, your checkout and your product catalogs. Seven weeks before Black Friday, that changes how I would plan a holiday launch.

The short version: Shopify Rollouts started as a way to schedule theme changes and run simple A/B tests. Since late September it can also switch discounts and market catalogs on and off, for everyone or for a share of visitors, on a timetable you set in advance, and roll the lot back automatically when the event ends. The catch sits on the developer side. A discount inside a rollout can be live for buyers while its own dates say it is expired, or the reverse, and apps that have not caught up can report it wrongly or lose sight of it. This article covers what Rollouts is, what changed, how far its A/B testing can be trusted, which apps to check, and the plan I would follow before the holiday traffic arrives.

A definition to anchor the rest: Shopify Rollouts is the native Shopify admin feature, found under Markets > Rollouts, that packages changes to a store’s theme, checkout and accounts pages, product catalogs and discounts so they can be scheduled, released to a percentage of visitors, run as a temporary event that reverts itself, or tested against the live store as an experiment.

In This Article

  1. What Shopify Announced on October 5
  2. What Shopify Rollouts Are: Launches, Events and Experiments
  3. How Rollouts Got Here: A Ten-Month Timeline
  4. Discounts and Catalogs in Rollouts: Promotions Now Move With the Store
  5. Shopify A/B Testing With Rollouts: What It Measures and How Much Traffic You Need
  6. Rollouts vs. A/B Testing Apps: What to Keep and What to Cancel
  7. Which Apps Can Get Your Discounts Wrong (and How to Check Yours)
  8. Plans, Permissions and Limits Before You Start
  9. A Quick Overview of Ruby Digital Agency
  10. A Ten-Step Plan for Using Rollouts Before Black Friday
  11. What Happens If You Do Nothing
  12. Frequently Asked Questions
  13. Final Thoughts

Key Takeaways

  • Shopify Rollouts lives under Markets > Rollouts and comes in three types: a Launch makes changes permanent, an Event runs them for a set period and then reverts, and an Experiment tests a treatment against a control.
  • Since September 2026 a rollout can carry discounts and market catalogs as well as theme and checkout-and-accounts changes, so a whole campaign can be scheduled and rolled back as one unit.
  • A discount inside a running rollout keeps its own start and end dates, so its status in the admin or in an app can contradict what buyers actually get. Shopify asked apps to read the new rollouts data on API version 2026-10.
  • Apps on older API versions silently lose sight of some rollout discounts. Discount, loyalty, reporting and marketing apps are the ones to ask about.
  • Experiments need the Grow plan or higher and real traffic: at a 3% conversion rate, detecting a 10% lift takes roughly 53,000 visitors in each arm.
  • Use Rollouts this season for every planned change, keep partial-traffic discounts away from codes you email to everyone, and keep a dedicated testing app only if you test prices or run a continuous testing program.

What Shopify Announced on October 5

The email arrived on October 5, 2026, from Shopify’s developer communications list. It tells app developers that merchants can now include discounts in rollouts. Every discount type in the Admin GraphQL API version 2026-10 gains a read-only rollouts connection, and apps should add a new access scope, read_rollouts, and read that data alongside a discount’s own fields. If they don’t, the email warns, they may receive incomplete discount data from merchants who use the feature.

The email is the last link in a short chain. Shopify posted two developer changelog entries dated October 1. The first, Discounts can now be included in Rollouts, is flagged as action required and lists September 16 as the date the change took effect. The second, Rollouts queries and webhooks in the Admin GraphQL API, lets apps list every rollout on a store and see what each one changes across discounts, catalogs, themes and checkout. Apps can also subscribe to notifications when a rollout starts, stops or changes its reach. A detailed upgrade guide for discount rollouts explains the edge cases.

Read as a merchant, the three documents say something simpler. Shopify has quietly finished turning Rollouts from a theme-scheduling tool into a campaign tool. You can now prepare a homepage, a checkout tweak, a market catalog and a set of discounts as one package, decide who sees it and when, and let Shopify undo it at the end. The developer notice exists because that package breaks an assumption almost every discount app has made for years: that a discount’s start and end dates tell you whether it is live.

The upgrade guide says it plainly. On version 2026-10, “a discount’s own fields don’t tell you what a rollout is doing to it.” Everything else in this article follows from that sentence.

DateDocumentWhat it means for a merchant
September 16, 2026Discounts in Rollouts takes effectDiscounts can be added to a rollout from the admin
October 1, 2026Two developer changelog entries, API 2026-10Apps get a way to see rollouts and their effect on discounts
October 5, 2026Email to app developersYour discount, loyalty and reporting apps have been told to update
November 27, 2026Black FridayThe first big sales weekend with discounts inside Rollouts

What Shopify Rollouts Are: Launches, Events and Experiments

Shopify’s help center describes a rollout as a scheduled set of changes to a store’s main theme, checkout and accounts pages, discounts, product catalogs, or any combination of them. You create one under Markets > Rollouts, give it a name, pick a type, add changes, then set an audience and dates. The type matters most, because it decides what happens when the rollout ends.

Rollout typeWhat it doesWhen it endsBest for
LaunchApplies changes to all eligible visitors, at once or graduallyChanges become permanentRedesigns, new checkout settings, planned go-lives
EventApplies changes for a fixed periodChanges roll back automaticallySales, holiday themes, flash promotions
ExperimentSplits traffic between a treatment and a controlChanges roll back; you choose a winnerTesting a theme, checkout or offer before committing

According to Shopify’s rollout types page, an Experiment defaults to a 50/50 split and a 90-day run, both of which you can change before saving. Launches and Events can also be given to fewer than 100% of visitors, which is how a gradual release works: show a new theme to 10% of traffic on Monday, watch for problems, and widen it later in the week. If a rollout reaches fewer than 100% of visitors, Shopify requires both a start date and an end date.

Five things are worth understanding before you click anything:

  • You edit a copy, not your live theme. When you add theme changes, Shopify works on a copy of your published theme, so you can keep editing the live store while the rollout waits. You can also swap in a completely different theme that you have already added to the store.
  • One theme change and one checkout change per rollout. The create a rollout page limits each rollout to one theme change and one checkout-and-accounts change. Catalogs and discounts can be added in numbers, up to 250 discounts per rollout.
  • A Launch can’t be undone once it ends. When a Launch reaches its scheduled time it applies the changes permanently and finishes. Shopify warns that there is no revert after that point.
  • Ending an Event early is a choice. If you end a running rollout yourself, Shopify asks whether to roll the changes back or apply them for good. Applying is permanent.
  • Archiving deletes theme work. The managing rollouts page notes that archived rollouts can’t be restarted, deleted ones can’t be recovered, and archiving permanently deletes the rollout’s theme changes. Export or duplicate anything you want to keep first.

The statuses are what you would expect: Draft, Scheduled, Active, Paused and Ended. A rollout can be paused and resumed, and an active rollout can still be edited. Editing either arm of a running Experiment, though, muddies its results, so treat a live test as frozen.

The mental model I give clients is a staging lane in the admin. For years, a holiday theme meant duplicating the theme, editing the copy, then someone staying up until midnight to press Publish, and someone else remembering to switch it back on December 1. Rollouts turns that into a dated, named object that does both jobs on its own, and now it can carry the discount that goes with the theme.

How Rollouts Got Here: A Ten-Month Timeline

Rollouts has grown in stages, and some of what has been written about it is already out of date. Several reviews from the spring say it cannot test checkout or offers. That was true in March and is not true now. Here is the sequence, from Shopify’s own announcements.

WhenWhat shippedSource
December 10, 2025Rollouts announced in the Winter ’26 Edition as native experimentation in the adminShopify Winter ’26 Edition
March 31, 2026Rollouts launches: schedule theme changes, auto-revert events, A/B test themesShopify changelog
June 5, 2026Scheduling, gradual publishing and A/B testing extended to checkout and customer accountsShopify changelog
September 16, 2026Discounts can be included in rolloutsDeveloper changelog
October 1, 2026API 2026-10: rollouts queries, six webhooks, rollouts on every discount typeDeveloper changelog
October 5, 2026Shopify emails app developers to updateShopify email

Market catalogs also appear in the help center as a change type, and the October 1 changelog lists catalogs among the resources an app can now see in a rollout. I could not find a dated merchant announcement for catalogs on its own, so treat October 2026 as the point at which it is documented rather than the day it arrived.

The Winter ’26 Edition framed Rollouts as a way to bring testing into the admin without outside tools. The March release delivered the theme half of that promise. The June release added checkout and customer accounts, which matters because checkout changes are the ones merchants are most nervous about making blind. The September and October changes add the commercial half: what you charge and what you discount.

Read the sequence and the direction is clear. Shopify wants every material change to a store to pass through a rollout: scheduled, staged, measurable and reversible. It fits the rest of the platform’s recent history. Shopify Scripts were retired in favor of Functions, and shipping is moving into Markets. The admin is becoming the single place where store logic lives, and Rollouts is how that logic changes over time. That is also why Rollouts sits under Markets rather than under Online Store. It is built on the same audience machinery that decides which buyers see which prices, catalogs and languages.

Discounts and Catalogs in Rollouts: Promotions Now Move With the Store

Until September, a Black Friday campaign on Shopify had two clocks. The theme changes had one schedule, set in Rollouts or by hand. The discounts had another, set on each discount’s start and end dates. Keeping the two in step was a spreadsheet job, and the classic failure was a homepage banner promising 30% off while the automatic discount behind it had not started, or had already ended.

Now a rollout can include changes that activate or deactivate existing discounts. You build the discounts as normal, then add them to the rollout. While the rollout is serving, a treatment can switch a discount on for the buyers it reaches, even if the discount’s own dates say it is scheduled or expired. It can also switch a discount off for those buyers, even if its own dates say it is live. When the rollout stops serving, the discount goes back to obeying its own dates.

The upgrade guide sets out what that means for the status you see. I have paraphrased its table:

What the rollout doesStatus shown on the discountWhat buyers actually get
Activates the discountScheduled or ExpiredBuyers in that treatment get it
Expires it for every buyerActiveNobody gets it
Expires it for some buyersActiveBuyers in that treatment don’t get it
Rollout is draft, scheduled, paused or endedAccurateWhatever the discount’s own dates say

Market catalogs work in a similar way. A rollout can set a market catalog to Active or Draft, so a catalog with its own product selection or price adjustments for one market can be switched on for a defined window. B2B and archived catalogs are excluded, and a catalog must be assigned to at least one market before its changes show. Status changes take effect immediately; edits to products and prices inside the catalog apply when the catalog is active.

Four practical rules follow from the requirements and considerations page and from how the feature behaves:

  • A discount can sit in only one unended rollout at a time. The same goes for a catalog. If you plan an early-access event and a main sale with the same discount, they have to run in sequence, not overlap.
  • Partial traffic means Online Store only. A discount activated at 100% effective traffic works on every sales channel, subject to its normal eligibility rules. A discount activated for part of your traffic applies on the Online Store only, so POS, marketplaces and other channels will not honor it.
  • Experiment arms can see different discounts. Combination rules still apply, and Shopify notes that treatment and control may have access to different sets of discounts. That is the point of a discount test, but it needs explaining to your support team.
  • Don’t split-test a code you have emailed to everyone. This one is my inference rather than a Shopify rule. If a code goes out to your whole list and a rollout only activates it for half your visitors, the other half type it in and get an error. Test automatic discounts, or codes sent only to the people who will see them.

The upside is significant. Earlier reviews of Rollouts, such as Convert’s analysis, criticized it for leaving out the commercial levers merchants care about most. With discounts and catalogs inside rollouts, you can now test whether a free-shipping discount beats 15% off, schedule a sale and the theme that advertises it as one object, or open a regional catalog for a holiday window. None of it needs an app.

Shopify A/B Testing With Rollouts: What It Measures and How Much Traffic You Need

The Experiment type is the part of Rollouts that gets the most attention, and the part I would approach with the most care. Shopify A/B testing in Rollouts is real, server-side and free on the Grow plan and above. It is also deliberately simple. Here is what Shopify’s analytics page says it reports:

Experiment containsMetrics reported
Theme changesConversion rate over time, bounce rate, reached-checkout rate, add-to-cart rate
Checkout changesCheckout conversion rate
Theme and checkout changesThe storefront metrics above, plus checkout conversion rate

What the help center does not document is the statistical method: how confidence is calculated, how visitors are assigned and kept in their arm, or how bots are excluded. Third-party guides disagree on what the report shows, with some describing confidence ranges and others saying significance indicators are missing. Shopify also notes that the same changes run as a Launch and as an Experiment can produce different numbers for the same metric, because the two are calculated through different pipelines. My advice is to treat an Experiment as a well-built comparison, not as a verdict, and to check its result against your own order data before you act on it.

The bigger constraint is traffic, and it applies to every testing tool, not just Shopify’s. To show how much traffic a reliable answer needs, I ran the standard sample-size calculation for a two-arm test (95% confidence, 80% power, two-sided). The figures are visitors needed in each arm:

Lift you want to detectBaseline conversion 2%Baseline conversion 3%
5% relative (e.g. 3.00% to 3.15%)about 315,000about 208,000
10% relative (e.g. 3.0% to 3.3%)about 81,000about 53,000
20% relative (e.g. 3.0% to 3.6%)about 21,000about 14,000
30% relative (e.g. 3.0% to 3.9%)about 9,800about 6,500

Put that against a real store. One with 50,000 sessions a month, split 50/50, puts about 25,000 visitors into each arm a month. It can detect a 20% swing in conversion in two to three weeks. A 10% swing takes about two months, and a 5% swing is out of reach. Most changes worth making move conversion by single-digit percentages. So on a mid-sized store, an Experiment is good at catching a change that is clearly worse, and poor at proving a small change is better.

That points to three sensible uses:

  • Safety checks. Before a new theme or checkout configuration goes to everyone, run it against the control for a fortnight. If add-to-cart or checkout conversion falls off a cliff, you have learned something valuable for free.
  • Big, bold comparisons. Two different homepages, two different offer structures, free shipping against a percentage discount. Large differences are detectable at moderate traffic.
  • Holiday-sized traffic. Black Friday week can deliver a month’s sessions in days. If you want to test an offer, that is when you have the volume. Decide in advance which result you will act on, and don’t peek and stop early.

What I would not do is run a string of small button-color tests on a store with 20,000 sessions a month and treat the winners as fact. That was poor practice with paid tools and it is poor practice with a free one.

Rollouts vs. A/B Testing Apps: What to Keep and What to Cancel

The question I am being asked most is whether Rollouts replaces the testing app a store already pays for. The honest answer is sometimes. Dedicated tools such as Intelligems and Shoplift charge anywhere from under a hundred to several hundred dollars a month. They earn that by doing things Rollouts doesn’t: price testing at the product level, audience targeting, custom goals, and statistics they explain. Rollouts earns its place by being free, built in, and able to schedule and revert as well as test.

NeedRollouts (October 2026)Dedicated testing app
Schedule a theme, checkout, catalog or discount changeYes, nativeUsually not
Auto-revert after an eventYes (Event type)Usually not
Gradual release to a % of visitorsYesSome
Whole-theme A/B testYes, Grow plan and aboveYes
Checkout configuration testYes, since June 2026Limited
Discount or offer testYes, since September 2026 (Online Store only below 100%)Yes
Product-level price testNot directlyYes, the main reason to keep one
Audience targetingMarket targeting onlyYes, often detailed
Custom goals and documented statisticsNoYes
CostIncluded in your planMonthly subscription

My rule of thumb for clients:

  • Cancel the app if you used it mainly for theme-against-theme tests or for scheduling, and you run a test or two a quarter. Rollouts covers that now.
  • Keep the app if you test prices, shipping thresholds or product-level offers, or if you run a continuous testing program with someone who reads the statistics. That person will want the targeting and the documented methods.
  • Don’t run both on the same pages at once. Two systems splitting the same visitors make both sets of results harder to read. If you keep an app, give it its own pages or its own weeks.

One caution on price testing. A market catalog with price adjustments can now be switched on for part of your traffic, which looks like a price test. Until Shopify documents it as one, I would treat that as a regional pricing tool rather than a way to test prices. Shoppers comparing notes, and marketplace feeds that only see the 100% version, make partial price changes riskier than partial theme changes.

Which Apps Can Get Your Discounts Wrong (and How to Check Yours)

This is the part of the October 5 email aimed at developers, but its effects land on merchants. Shopify’s upgrade guide sorts apps by what they do with discounts and says what each type must change. Here is my plain-English version, with the symptom you would see if an app has not updated:

If an app…Typical examplesSymptom if it hasn’t updated
Caches or syncs discountsLoyalty, ERP, marketplace connectorsKeeps showing a discount as live or expired after a rollout changes it
Shows discounts to you in its own screensPromotion managers, bundle appsStatus contradicts what customers actually get
Promotes discounts to shoppersBanner and popup apps, sales channels, email toolsAdvertises an offer that only part of your traffic can use
Acts on a discount’s statusAutomation and workflow appsFires on the wrong state
Reports discount performanceAnalytics and attribution toolsCredits a 50%-traffic discount with store-wide results
Deletes records when a discount disappearsSync and backup toolsTreats a hidden discount as deleted
Only creates or writes discountsDiscount generatorsNo change needed

The last two rows deserve explanation, because they show how a quiet API change can do real damage. Apps still on an API version earlier than 2026-10 cannot represent a rollout. So Shopify leaves those discounts out of their results whenever the rollout makes the discount’s status misleading. The discount disappears from lookups, lists and counts, and nothing in the response says why. It comes back when the rollout stops affecting it. An app that deletes its own copy of anything that vanishes could delete records it should have kept. Shopify’s guidance to developers is blunt: an absent discount is not a deleted one.

The fix on the developer side is clear. Move to 2026-10, request the read_rollouts scope, read the rollouts on each discount, and subscribe to the six new rollout webhooks so the app hears when a rollout starts, stops or changes its reach. Because read_rollouts is a new permission, installed apps may need you, the merchant, to approve it. Expect some apps to ask you to reauthorize them over the coming weeks. That is normal and you should approve it.

There is one more subtlety. A rollout has a configured traffic allocation, the percentage you set, and an effective one, which Shopify calculates after resolving conflicts with other active rollouts on the same resources. The effective figure can be lower. Shopify’s Rollout API reference exposes both. An app that wants to be sure a discount reaches everyone should check that the effective allocation is 100% and that a single treatment activates it at a 100% split. You don’t need to know the field names, but you should know the question to ask your app vendors.

What to do as a merchant takes about an hour:

  • List the apps that touch discounts. In the admin, open Settings > Apps and note every app that reads, shows, promotes, reports on or syncs discounts. Loyalty, bundles, upsells, email and SMS, marketplace and ERP connectors, and analytics are the usual suspects.
  • Ask one question. Email each vendor: do you read discount rollouts on Admin API 2026-10, and when will you request read_rollouts? A vendor who knows the answer immediately is one you can trust with Black Friday.
  • Keep rollout discounts away from apps that say “not yet”. Until a vendor confirms, schedule the discounts those apps depend on with their own dates rather than inside a rollout.

Plans, Permissions and Limits Before You Start

Rollouts is broadly available, but a few features are gated by plan and a few setups are not supported at all. Here is the summary from Shopify’s requirements page and the original March announcement:

ItemRequirement or limit
Rollouts (Launch and Event)Basic plan or higher
Experiments (A/B tests)Grow plan or higher
Market-specific targetingAdvanced or Plus (per the March 2026 announcement)
Discounts per rolloutUp to 250
One discount or catalogIn only one unended rollout at a time
Theme and checkout changesOne of each per rollout
Vintage themesNot supported
Headless or custom storefront checkoutsNot supported for theme or checkout-and-accounts changes
Experiments and partial trafficOnline Store only

On permissions: the store owner and administrators can manage rollouts. Staff need the Markets > Rollouts permission, plus the matching permission for each kind of change: checkout and customer accounts, Discounts, Online Store themes (and Edit code to change the current theme), and Catalogs view or edit. If you have an agency or freelancer as a collaborator, check those permissions before the week you need them.

The help center also lists Liquid templates among the things a rollout can’t carry. In practice, if a change needs new code rather than theme-editor settings, build it in a separate theme and use a rollout to replace the main theme with it. Do that well before a launch, because a full replacement theme is a bigger change to test than a set of editor tweaks.

Two reach rules are easy to miss. First, when a rollout serves fewer than 100% of visitors, Shopify’s reach and external channels page says external channels receive no rollout information, so ads, feeds and listings keep showing the live store. Second, at 100%, changes do flow to the Storefront API, POS and product feeds such as Google and YouTube, but connected channels update on their own schedules. A catalog or price change can take time to appear everywhere, so start Events that touch products a little early.

On SEO, the same page says organic search results are not put at risk by a rollout as such. Product availability changes, URL changes and page errors still affect indexed pages, as they always have. A theme experiment that breaks a template is still a broken template to Google.

A Quick Overview of Ruby Digital Agency (and Why We Care)

For readers meeting us for the first time: Ruby Digital Agency is a Shopify Plus Partner that does one thing on purpose. We move brands onto Shopify and Shopify Plus and make the move pay for itself. Most of our clients arrive from BigCommerce, WooCommerce, Magento and Adobe Commerce, OpenCart, Shift4Shop, PrestaShop or a custom build, usually because their platform has become expensive to run, hard to change or both. We are a boutique team: the specialists who plan a project are the people who deliver it, and every migration is run on our own AI-driven migration technology, with every record checked by a person against the source store before go-live.

ServiceWho it is forTypical timeline
Standard Shopify MigrationMid-market brands, $2M–$20M/yr4–8 weeks
Shopify Plus ReplatformingEnterprise brands, $20M+/yr, from Magento, Adobe Commerce or custom builds6–8 weeks
Multi-Store & Subscription IntegrationB2B wholesale, subscriptions, multi-brand architecturesScoped per project
Ecommerce Website AuditsAny platform: Shopify, WooCommerce, BigCommerce, Magento, customFixed-scope engagement
SEO and AEOBrands that need to be found in Google and cited by AI assistantsOngoing
Shopify POS ImplementationRetailers uniting stores and onlineScoped per project

The economics are why brands call us. Across our migrations, clients recover the investment in under 18 months, see conversion rates about 15% higher on average, and cut total cost of ownership by around 29%; infrastructure costs alone fall by as much as 80% when a self-hosted stack goes away.

Rollouts matters to us for a simple reason. A migration ends with a launch, and the first few months after a launch are when a merchant changes the most: new homepage, new collection structure, new offers, the first holiday season on a new platform. Many of the stores we move come from platforms where every one of those changes went live the moment someone pressed save. Teaching a team to stage, schedule and roll back their changes is now part of how we hand a store over. We also build custom apps for clients, and the October 5 notice puts any of them that read discounts on our own update list. If you would like a second pair of eyes before the holidays, our ecommerce website audits now include a rollout and app-readiness review. We check your plan for the season, the apps that touch your discounts, and whether your traffic can support the tests you want to run.

A Ten-Step Plan for Using Rollouts Before Black Friday

Black Friday falls on November 27 this year, seven and a half weeks from today. This is the order I would work in. Most stores can finish the first four steps this week.

  1. Check your plan and permissions. Confirm whether you are on Grow or higher if you want to experiment, and give the staff and collaborators who will build the holiday rollouts the Markets > Rollouts permission plus the change-specific ones. A missing permission discovered on November 26 is a bad evening.
  2. Audit the apps that touch discounts. List them and ask each vendor about API 2026-10 and read_rollouts. Approve any reauthorization requests that follow. Note which vendors have not committed to a date.
  3. Map the season as rollouts, not tasks. Write down each event with its start and end: early access, Black Friday, Cyber Monday, December shipping cut-off, January sale. Remember that one discount or catalog can sit in only one unended rollout at a time, so plan events in sequence.
  4. Build the discounts first, in draft. Create every discount with sensible fallback dates of its own. If a rollout is paused or ended early, the discount goes back to those dates, so they should never cause harm.
  5. Use Events for anything temporary. Holiday theme, sale banners, the Black Friday discount set and any regional catalog belong in an Event with an end date, so the store reverts itself while you sleep.
  6. Keep emailed codes at 100%. Any code sent to your whole list or posted publicly should be in a 100% rollout or outside rollouts altogether. Reserve partial-traffic discounts for automatic discounts or targeted sends.
  7. Run one safety experiment now, not in November. If the holiday theme or a checkout change is significant, test it against the control for two weeks in October. You are looking for a clear drop, not a small win.
  8. Decide your one holiday test in advance. If you want to test an offer during peak traffic, choose it now, decide the metric and the length now, and leave it alone until the end date.
  9. Rehearse the start and the end. Schedule a small dummy Event for an hour on a quiet day and watch it start and revert. Check your feeds, POS and connected channels during the window, given that they update on their own schedules.
  10. Write the runbook. One page: each rollout’s name, type, dates, traffic and discounts, who can pause it, and what “pause” does to the discounts inside it. Give a copy to customer service so they can explain why one shopper sees an offer and another does not.

Want a second pair of eyes on your holiday rollouts?

We review your planned rollouts, the apps that read your discounts, and whether your traffic can support the experiments you have in mind, then hand you a dated plan for the season. It is part of our ecommerce website audits. Book a call and we will look at your store before Black Friday.

What Happens If You Do Nothing

If you never open Markets > Rollouts, nothing breaks. Discounts outside rollouts behave exactly as they always have, and the API change only affects discounts you choose to put inside a rollout. That is the reassuring part.

The cost of ignoring Rollouts is the old way of running a season. Somebody publishes the holiday theme by hand at midnight. The discount and the banner run on separate clocks and drift apart. A December 1 revert gets forgotten until a customer points it out. Changes go to 100% of visitors untested because testing needed an app nobody bought. None of it is a disaster, but all of it costs money and sleep, and Rollouts now removes most of it at no extra charge.

The cost of half-adopting it is more specific. Put discounts inside rollouts without checking your apps, and you can get a loyalty program showing the wrong offers, a popup advertising a deal half your visitors can’t use, or a sync tool that treats a hidden discount as deleted. Those are the problems the October 5 email was written to prevent, and they are the ones that turn into support tickets on the busiest weekend of the year.

Frequently Asked Questions

What are Shopify Rollouts?

Shopify Rollouts is a native admin feature under Markets then Rollouts that packages changes to your theme, checkout and accounts pages, market catalogs and discounts. Each rollout can be a Launch that makes changes permanent, an Event that runs for a set period and then reverts, or an Experiment that tests a treatment against a control. You can schedule it, release it to a percentage of visitors and pause it.

Which Shopify plans include Rollouts?

Rollouts is available on the Basic plan and higher. Experiments, which are the A/B testing part of Rollouts, need the Grow plan or higher. When Rollouts launched in March 2026, Shopify said targeting a rollout at a specific market needs the Advanced or Plus plan. Vintage themes are not supported, and headless or custom storefront checkouts cannot take theme or checkout changes.

Can I add discounts to a Shopify rollout?

Yes. Since September 2026 a rollout can activate or deactivate existing discounts, up to 250 per rollout. While the rollout is serving, it overrides the discount dates for the buyers it reaches. When it stops, the discount follows its own dates again. A discount can be in only one unended rollout at a time, and below 100 percent traffic it applies on the Online Store only.

Why does my discount show as expired when customers can still use it?

It is probably inside a running rollout. A rollout does not change the start and end dates of a discount, so the status shown on the discount still comes from those dates. A treatment can activate a discount that looks scheduled or expired, or switch off one that looks active. Open Markets then Rollouts to see which rollout includes it and what each treatment does.

Can Shopify Rollouts replace my A/B testing app?

For whole-theme tests, checkout configuration tests, offer tests and scheduling, often yes, and it is included in your plan. Keep a dedicated testing app if you test product prices or shipping thresholds, need detailed audience targeting or custom goals, or run a continuous testing program where someone reads the statistics. Avoid running both tools on the same pages at the same time.

How much traffic does a Shopify A/B test need?

More than most stores expect. At a 3 percent conversion rate, detecting a 10 percent relative lift with 95 percent confidence and 80 percent power takes about 53,000 visitors in each arm. A 20 percent lift takes about 14,000. Smaller stores should use experiments to catch clearly worse changes or compare bold alternatives, and run them for at least two full weeks.

Do my apps need updating for discount rollouts?

Apps that cache, display, promote, report on or act on discounts should move to Admin API version 2026-10 and request the read_rollouts scope. On older versions, Shopify hides some rollout discounts from them without explanation. Apps that only create discounts need no change. Ask your loyalty, bundle, marketing, analytics and sync vendors for their update date, and approve any reauthorization request.

Do Shopify Rollouts affect SEO or Google Shopping?

Shopify says a rollout does not put organic listings at risk by itself. Below 100 percent traffic, external channels such as ads and product feeds keep seeing the live store. At 100 percent, changes flow to product feeds and other channels, but those update on their own schedules. Broken templates, URL changes and availability changes still affect indexed pages as usual.

Final Thoughts

The October 5 email was written for developers, but the change behind it is for merchants. Shopify Rollouts is now the one place in the admin where a whole campaign can be prepared, scheduled, tested and undone: theme, checkout, catalog and discounts together. That is a better way to run a season than the midnight-publish, separate-clocks routine most stores still follow, and it costs nothing extra.

My position for clients is simple. Use Rollouts for every planned change this holiday season. Use Events for everything temporary, so the store cleans up after itself. Treat Experiments as a safety check unless you have Black Friday volumes. And before you put a single discount inside a rollout, make sure every app that reads your discounts has caught up with API 2026-10.

Shopify has given merchants the scheduling and testing discipline that only well-run engineering teams used to have. The stores that benefit will be the ones that adopt the discipline along with the tool.

Planning a Move to Shopify or Shopify Plus?

We plan the strategy, run the heavy lifting with AI, verify every record against your source data before go-live, and hand over a store your team can change safely, rollouts included. Book a migration strategy call and we will map your move.

Related Reading

Lance Allison
Founder & CEO, Ruby Digital Agency

Lance Allison is a Shopify Plus Partner and eCommerce migration specialist based in Salt Lake City, Utah. He has helped hundreds of merchants move to Shopify and Shopify Plus, and specializes in complex, data-heavy platform migrations, and in the launch practices that keep a new store stable through its first busy season.

Keep reading

More from the journal.

Browse all

Ready when you are.

Tell us about your store in ten minutes. We come back within one business day with an honest assessment, a recommended track and a fixed price.