A discount is scheduled in a Shopify rollout. Is the stock ready?

Since October 2026 a Shopify rollout can include a discount. The marketing side of the business creates the rollout under Markets > Rollouts, picks a start date three weeks out, attaches the discount and the new theme banner, and schedules it. Everything about the promotion is now decided and written down in one place in the admin. Almost everything: nobody has yet asked whether the products in the discount will still be on the shelf on day two.

That gap is the subject of this post. A scheduled rollout is the best demand forecast most small retailers will ever get, because it has a date, a duration, a list of products and a rough idea of how many visitors it reaches. This post explains what the rollout tells you, how to turn it into a purchase order in Restock Ops with enough lead time, and why the app itself needed no change for Shopify's new feature.

What a rollout tells you that a normal sale does not

Shopify's overview of rollouts describes three kinds. A Launch merges its changes into the store permanently. An Event publishes them for a set period and rolls them back when it ends. An Experiment serves a treatment to a share of visitors and compares it with a control. All three can carry a discount change, which, according to Shopify's page on rollout types, activates or deactivates discounts you have already created for the visitors in the rollout.

Each of those settings is a planning input. The start date is when demand changes. An Event's end date is when it changes back, which matters as much: stock that arrives after the Event rolls back is stock you will carry at full price for a long time. The traffic allocation tells you what share of visitors see the offer at all, and an Experiment halves that again, since only the treatment group gets the discount. A 20% discount in an Experiment at 50% allocation reaches a quarter of your online store visitors, not all of them, and your order should reflect that.

One more detail from Shopify's requirements page: Experiments, and any rollout below 100% traffic, apply only to the online store. Point of sale is outside them. If most of your sales happen in the shop, a rollout experiment will barely move your stock at all, and the lift you plan for should be an online lift only.

Why the reorder report will not see it coming

The reorder report in Restock Ops suggests quantities from recent sales. For each variant it works out units sold per day over a recent window from your orders, multiplies that velocity by your supplier's lead time plus the days of cover you want, and compares the result with what you have on hand and on order. That is the right way to reorder a steady seller, and the wrong way to prepare for a promotion that has not started yet, because the window only contains the past.

This is not a flaw to fix by making the app read rollouts. The lift a discount produces depends on the product, the depth of the discount, the audience share and what else you are doing that week, and no formula in an inventory app knows those things better than you do. What the app can do is make it easy to override the velocity for the products in the rollout, for the length of the rollout, and then let the normal calculation take over again.

Every row on the reorder report has a manual velocity field. Leave it empty and the row uses the measured rate from orders. Enter a number and the suggestion uses that instead, with the same lead time and days of cover. That is where a rollout's forecast goes. Our guide to holiday ordering with order-by dates covers the same mechanics for a season rather than a single promotion.

Turn the rollout into a purchase order

1. Open the rollout in Markets > Rollouts and write down four things: the products in the discount change, the start date, the end date if it is an Event, and the traffic allocation. Note whether it is an Experiment, in which case roughly half the allocated traffic gets the discount.
2. Pick a lift. Look at the last comparable promotion in Shopify's reports and compare units per day during it with the weeks before. If you have never run one, a modest guess scaled by the audience share is better than nothing: a 40% lift at 100% allocation becomes a 10% lift at 25% reach.
3. In the Restock Ops reorder report, filter to the discounted products. For each one, set the manual velocity to the current rate times your lift, and check that days of cover at least spans the rollout. Save and read the suggested quantities.
4. Compare the suggestion against the supplier's lead time and the rollout's start date. If the order cannot arrive before the start, order what will arrive during the rollout and no more; stock that lands after an Event ends is next season's problem.
5. Create the purchase order from the report and send it. When the goods arrive, receive them in Restock Ops so Shopify's quantities are correct on the morning the rollout starts; the partial receiving guide covers deliveries that come in two shipments.
6. The day after the rollout ends, clear the manual velocity on those rows so the report goes back to measured sales, which will now include the promotion week and over-state demand for a while. Our post on holiday ordering explains how to read the report in the weeks after a spike.

What the discount's own dates no longer tell you

Before rollouts, the start and end dates on a discount were the only schedule it had. Now a discount keeps those dates and a rollout can switch it on or off independently for its audience. The discount is active for a customer when its own dates apply or when an active rollout that activates it covers them, and inactive for the customers in a rollout that deactivates it. In practice that means the Discounts page can show a discount as Scheduled while it is already live for half your online visitors, and you cannot plan stock from the Discounts page alone. Read the rollout.

Shopify made the same point to app developers in October 2026. From API version 2026-10 every discount exposes the rollouts that include it, and apps that read discount data were asked to request a read_rollouts permission and account for rollout schedules and traffic allocation. Restock Ops reads inventory, orders, purchase orders, transfers and locations. It never reads discounts, so the notice does not apply to it: there is no new permission to grant and nothing in the app changed. The planning work in this post is human work, and it would be whichever app you used.

Two mistakes to avoid

• Ordering for a 100% sale when the rollout is an Experiment. The order should match the audience that gets the discount, not the store's total traffic. Check the type and the allocation before you set a lift.
• Letting a promotion week poison the next reorder. After the rollout, measured velocity includes the spike. Clear the manual velocity, then be sceptical of suggestions for the discounted products until the window has moved past the promotion.

FAQ