Shopify Subscriptions Explained: How Recurring Orders Actually Run
First, check the fulfillment status
Everything about this request depends on one fact: has the order shipped yet. If it is unfulfilled, you have options. If it is fulfilled or in transit, you are no longer editing an order, you are arranging an exchange, and the conversation with the customer changes completely. Check the order timeline before you promise anything, because the wrong promise here creates the refund dispute you were trying to avoid.
In the Shopify admin, open the order and look at the fulfillment status first, the payment status second. An unfulfilled, paid order is the easy case. A partially fulfilled order means some items can still change and some cannot, which you need to explain clearly. Train whoever answers customer messages to check this before typing a reply; most bad outcomes in this scenario come from promising an edit that was never possible.
When you can edit the order
For unfulfilled orders, Shopify lets you edit: add or remove line items, change quantities, and adjust the order total, with the customer charged or refunded the difference. This is the cleanest path when the customer wants a different size or color of the same product. Make the edit, confirm the new total with the customer in writing, and note that the edit itself is the confirmation; do not also send a separate "we changed it" email that confuses the thread.
Edits have limits worth knowing. You cannot change the order into something unrecognizable without confusing your own records, and editing does not work once fulfillment has started. If the price difference is significant, get explicit approval before charging the extra amount; an unexpected charge on a card is how a simple size swap becomes a chargeback.
When to cancel and reorder instead
Cancel and reorder is the right call when the change is extensive (half the items are different), when the customer wants to add items that change shipping weight or method, or when the original order is already fulfilled. It is also cleaner when the customer is clearly unsure; a fresh order with a fresh confirmation beats three rounds of edits that nobody can follow.
The mechanics matter: cancel the original, issue the refund promptly, and create the new order, ideally as a draft order you send for the customer to approve and pay. Do not cancel without the replacement ready, or the customer experiences it as "my order was cancelled" rather than "my order was fixed." Speed of the refund is the whole ballgame here; a same-day refund turns a hassle into a loyalty moment.
What to tell the customer
The message template that works: confirm what you are doing, state the timeline, and name the money. "I have updated your order to the blue in size medium. The $12 difference has been refunded and should appear in 3 to 5 business days. Your order still ships Thursday." Every element answers a question the customer would otherwise ask in a follow-up message, which is the real cost driver in these exchanges.
If you cannot do what they asked, say so in the first sentence and offer the alternative immediately. "That order already shipped this morning, so I cannot change it, but here is what I can do: refuse the delivery or send it back with this prepaid label, and I will ship the correct item today." Bad news first, solution second, no hedging in between.
Preventing it next time
A sizing chart linked at the point of size selection, a clear "final sale" flag on items that cannot be changed, and an order confirmation email that arrives within minutes all reduce these requests. The confirmation email is the cheapest intervention: customers who catch their own mistake within the hour are easy to help, while customers who notice three days later are not.
Track how often this happens and why. If size swaps dominate, your size guide is failing. If customers keep adding forgotten items, your checkout may be rushing them. The pattern in the requests is free product-page research, and merchants who read it stop answering the same email every day.
A subscription is a selling plan plus an app that keeps promises
Shopify itself provides the framework: selling plans define the offer, such as delivery every 30 days at 10 percent off, and the contract that records what the customer agreed to. The subscription app, Shopify own Subscriptions app or a third party, provides the machinery around it: the checkout option, the customer portal, the renewal scheduling, the retry emails. Understanding which layer does what saves hours when something breaks, because plan problems and app problems have different fixes and different owners.
At checkout the customer buys the first order normally and consents to the recurring terms: amount, frequency, and the cancellation route. That consent record matters. It is what separates a subscription from a surprise, in card network rules and in the customer head, and the stores that bury it buy chargebacks with their retention rate.
Renewal orders are new orders, with new ways to fail
Each cycle, the app creates a fresh order and attempts payment on the stored method. Failures are routine and mostly innocent: expired cards, replaced cards, insufficient funds on the wrong day. Dunning is the recovery system, a short sequence of retries on a schedule plus emails asking the customer to update payment in the portal. A store without dunning configured is donating a percentage of its subscriber base to card expiry every month, silently.
Inventory and pricing at renewal follow the contract and your settings. A product that goes out of stock does not stop the calendar, it creates a failed or partial renewal someone must handle. Decide the rules before launch: skip, substitute, or cancel on stockout, what happens when your price rises, and how discounts behave after the first order. Subscribers forgive a lot except finding out from a bank statement.
The portal is the product, for retention purposes
Customers judge a subscription by how it feels to control. The portal needs to make the five actions effortless: skip the next order, swap a product, change frequency, update payment, and cancel. Every one of those that requires an email to support converts control frustration into cancellations and, worse, into disputes. Paradoxically, an easy cancel button retains better than a maze: subscribers who can pause stay subscribers, while trapped subscribers cancel permanently and tell people why.
Cancellation reasons collected at that moment are the cheapest research you will ever buy. Too much product is a frequency fix. Price is a tier or a loyalty fix. Quality is the one that should reach the founder desk unfiltered.
Run the churn math before the launch party
Subscriptions look like free recurring revenue until you model them: a subscriber acquired at a discount, churning after two cycles, often returns less than the same customer bought one time at full price, and the support and dunning overhead is real. The categories that win are consumables with a natural use rate, coffee, pet food, supplements, household refills, where the subscription solves a genuine running out problem on a predictable clock.
Track cohort retention by cycle, not a blended subscriber count, because growth hides rot: new signups can mask a base that leaves at cycle three. The blended number is the one in the pitch deck. The cohort table is the business.
Related reading
If this topic is useful, these earlier guides go deeper on connected questions: Shopify Payment Pending Explained: Why an Order Shows as Not Paid Yet, Shopify Order Status Page Explained: What Customers Actually See, and Shopify Markets Explained: One Store, Several Countries, Real Limits.
Frequently asked questions
Do I need an app for Shopify subscriptions?
You need subscription functionality beyond the base platform: Shopify offers its own free Subscriptions app, and third party apps add deeper portals, dunning, and analytics. Shopify provides the selling plan and contract framework underneath whichever app you choose.
What happens when a subscription payment fails?
The renewal order fails and the app dunning sequence takes over: scheduled retries plus emails asking the customer to update their payment method in the portal. Without dunning configured, failed renewals quietly become lost subscribers, so set retries and notifications before launch.
Can customers pause or cancel easily?
They should be able to from the customer portal: skip, pause, swap, change frequency, update payment, and cancel. Making cancellation difficult tends to increase permanent churn and payment disputes, while easy control, especially pause and skip, keeps more subscribers long term.
Are subscriptions profitable for every store?
No. They suit consumables with predictable use rates, where reordering is a real chore for the customer. Model cohort retention honestly: a discounted subscriber who leaves after two cycles can be worth less than a one time full price buyer, before counting support overhead.