What is subscription ecommerce?
Subscription ecommerce is an online selling model in which a customer agrees once to receive a product on a schedule and the store charges a saved payment method each cycle without asking again. The purchase decision is made one time. Every later order is a charge against that decision, and the store’s job changes from earning each sale to keeping the agreement alive.
Most definitions stop there and list three variants. That is accurate and not very useful. The useful version is the operator’s: a subscription is a liability and an asset at the same time. It is a promise to ship product in the future, and it is a forecastable stream of cash. Both halves need someone running them.
This article is written for operators at roughly $3M to $30M in revenue, on Shopify Plus or another paid platform. If you are below the $3M floor we publish, the mechanics still apply, but the tooling and analytics work will not pay for itself yet, and this is not written to you. If your product is bought once or twice in its lifetime, subscription is probably the wrong model, and the section on fit explains why.
The three parts of the definition
A working subscription has three parts, and each one can fail alone.
- The agreement. The customer chose a product, a frequency and a price, and consented to be charged repeatedly.
- The billing engine. Something holds the schedule, asks the payment gateway to charge a stored token, and creates the order in Shopify when payment clears.
- The fulfilment promise. A warehouse or 3PL has to hold enough stock, on the right date, for people who have not placed an order in the usual sense.
A dictionary would call this “recurring commerce”. An operator calls it three systems that must agree with each other every night.
What does subscription ecommerce change for an operator?
Subscription ecommerce changes four things: how revenue is counted, how stock is planned, which metric decides profit, and who in the business owns the customer after the first order. Marketing usually gets the credit for a subscription launch. Finance, operations and support get the consequences.
Revenue becomes a ledger of commitments
A one-off store books revenue when an order clears. A subscription store also carries a book of future charges, each with a date, an amount and a probability of succeeding. That is closer to a receivables ledger than to a sales log.
The practical effect is that two numbers that matched in a one-off store now diverge. Bookings (subscribers times price) and cash collected (charges that actually cleared) drift apart every cycle, and the gap is your leak. Baremetrics puts the loss from failed payments at about 9% of MRR, an independent figure drawn from subscription businesses generally, not from ecommerce alone.
Stock is bought against a schedule, not a forecast
Retailers plan stock from demand estimates. A subscription store has a better input: a list of contracts with next-ship dates. That is a gift, as long as you use it.
Read the schedule before you place a purchase order. Contracts due to ship in the next six weeks are, in effect, pre-sold units. Paused and skipped contracts are the adjustment. Teams that plan from historic sales alone end up either short on the biggest renewal day of the month or holding stock for subscribers who have already left. If you run purchase order automation, feed it contract dates, not last quarter’s order history.
Churn replaces conversion as the number that matters
In a one-off store, conversion rate and average order value decide profit. In a subscription store, the survival curve of each signup cohort does. A small change in monthly retention compounds across every future shipment.
Cohort retention is why the same customer-acquisition cost can be sensible for one brand and ruinous for another. What the payback depends on is how many shipments the average subscriber receives before cancelling. LTV to CAC is the ratio to watch, and its denominator only makes sense once you know your own cohort retention. Industry averages are a starting sanity check; the average churn rate for subscription services article covers how to read them without borrowing someone else’s number.
The customer relationship moves from marketing to operations
After the first order, a subscriber’s experience is a run of transactional events: a renewal reminder, a charge, a shipping notice, a delivery. Nobody sends a campaign. Yet every one of those events is a chance to lose the customer.
Support tickets shift too. Expect “I didn’t realise I’d be charged again”, “my card changed” and “I have too much of this”. Those are billing and fit problems. A discount code does not solve any of them.
Which models sit inside subscription ecommerce?
Three models cover almost every store, and they break in different places. Choosing the wrong one for your product is the single most common design error.
| Model | What the customer agreed to | Where it breaks first |
|---|---|---|
| Replenishment | The same consumable on a repeating schedule | Usage rate varies, so stock piles up and the customer cancels around the third shipment |
| Curated box | A changing selection chosen by the brand | Cost of goods drifts from the price promised, and fulfilment gets complicated by variety |
| Membership or access | A recurring fee for pricing, shipping or early access | The benefit is not felt between purchases, so the fee reads as an unused charge |
Replenishment is the easiest to run and the easiest to get wrong, because the product’s real consumption speed is set by the customer, not by you. A subscription box adds a merchandising problem on top of the billing problem. Memberships depend on a benefit the member can feel; if it only shows up at checkout, it tends to look like a fee for nothing.
Product fit gets tested at the third shipment more than at signup. Our article on why supplement subscribers cancel at month three walks through one version of that pattern.
How does subscription ecommerce work on Shopify?
Subscription ecommerce on Shopify works through a subscription app that sits alongside the store. Shopify supports selling a product with a recurring option through its Selling Plans, and the app you install manages the schedule, the customer portal and the renewal charges. Shopify itself is not the subscription engine.
The flow is the same across apps:
- A customer picks a selling plan on the product page, for example “every four weeks”.
- At checkout their payment method is saved with the payment gateway.
- On each renewal date the app asks the gateway to charge the saved method.
- When the charge clears, the app creates a normal Shopify order, which your fulfilment stack picks up.
The two systems hold different halves of the truth. Shopify has the orders. The app has the contracts, including the paused, skipped and failed ones that have not produced an order. A report built only on Shopify orders will never show you a subscriber who is at risk.
Apps in this space include Recharge, Skio, Loop and Stay AI, and each packages its pricing differently. Check current plan terms on each vendor’s own pricing page. For comparisons, see best subscription app for Shopify, Loop Subscriptions and Recharge alternatives. Those cover selection; this article is about what the model does to your business once one is chosen.
A specific Shopify Plus point applies. Checkout customisation on Plus can change how the subscription offer is presented and how cross-sells are added to a recurring order, and API limits are higher at volume. Neither is required to start. Confirm with your app’s documentation which of your Shopify plan’s features it uses.
Where do operators go wrong with subscription ecommerce?
Operators go wrong in five repeatable ways, and none of them is a marketing mistake. Every one is a problem of running the ledger.
They treat signups as the success metric
A signup dashboard rewards the discount that brought the customer in. It says nothing about whether the customer stayed. Track survival by cohort, meaning the share of each month’s signups still active in months two, three and six. If you do not have that view, build it before you spend on acquisition.
They ignore involuntary churn
Some cancellations are not decisions. A card expires, a bank flags the charge, or funds are short on the renewal date. The customer never chose to leave. Paddle and ProfitWell put involuntary churn at 20 to 40% of all churn, and Stripe attributes roughly 25% of lapsed subscriptions to payment failure. Both are vendor-measured across their own customer bases, so read them as a range for your case rather than a forecast.
Here is what that looks like in a hypothetical case. As a hypothetical, suppose a store has 2,000 active subscribers on a $40 monthly order. Say 100 charges decline in a given cycle, and the retry sequence recovers 60 of them. The 40 that do not recover represent $1,600 of billings that cycle, from customers who never said they wanted to go. The fix is not a new offer. It is a better retry schedule, a card-updater, and a message sent while the customer is still willing. See involuntary churn, how to reduce involuntary churn and dunning management for the mechanics.
They hide the exit
A cancel button buried three screens deep does not reduce churn. It moves the cancellation to the bank as a chargeback and to your inbox as an angry email. Put skip, pause, change frequency and swap product ahead of cancel, and keep cancel reachable. Rules on how easy cancellation must be vary by market, so confirm specifics with counsel.
They discount without a control
A first-order discount fills the funnel, but whether it fills with customers who stay is a separate question. Compare a discounted cohort with a full-price one across the same number of shipments, and judge on retained margin per subscriber. A discount that raises signups and lowers month-three survival can be a net loss.
They report one system
Shopify shows orders. The subscription app shows contracts. Finance wants cash. If those three do not reconcile, every number you produce is arguable. Pick one owner for the reconciliation and run it monthly. This is unglamorous work, and it decides whether you can believe your churn figure at all.
What is subscription ecommerce confused with?
People confuse subscription ecommerce with four neighbouring ideas. Each has a different commitment, and so a different failure mode.
- Repeat purchase. The customer buys again by choice. There is no stored agreement, no scheduled charge and no involuntary churn, but also no predictability.
- Auto-replenish reminders. The store emails when the customer is likely running low. The customer still decides each time. It lowers commitment and removes billing risk together.
- Loyalty programmes. Points and tiers reward purchases the customer would have made. They add no schedule. For that design problem, see ecommerce loyalty platform.
- SaaS subscriptions. The billing engine looks alike, but there is no physical cost per period. A SaaS cancellation stops revenue and cost together. A subscription ecommerce cancellation can strand stock you bought for the schedule.
The boundary matters because tools and metrics do not transfer. A SaaS churn benchmark applied to a replenishment brand gives a wrong answer with a confident label.
How do you know if your subscription programme is healthy?
You know a subscription programme is healthy by measuring cohort survival, involuntary churn and retained margin on your own data, then comparing them with a benchmark for your category. The order matters: your numbers first, someone else’s second.
Start with two steps.
- Work out your churn from cohorts. Take each signup month and count how many are still active at each later month. Separate voluntary cancellations from failed-payment lapses, because they need different fixes. The subscription churn calculator does the arithmetic once you have the counts.
- Compare with a category benchmark. For consumable DTC brands, the subscription churn benchmark for DTC consumables gives a reference point with its sources labelled. Treat it as a sanity check, not a target.
A figure that you cannot compute from your own contracts is marked metric to confirm until you can. That applies to ours as well. We do not publish an average churn rate for your category from our own accounts, and any article that gives you one without a source is guessing.
Two more checks are cheap. First, compare contracts in the subscription app against active customers in your email platform; a mismatch means a sync fault. Second, ask your 3PL for the next four weeks of scheduled ships and compare that with what is on the shelf. If those two conversations are uncomfortable, the model is running on hope. See 3PL fulfillment for what a warehouse needs from you to plan around a schedule.
Is subscription ecommerce a retention problem?
Subscription ecommerce is a retention problem first and an acquisition problem second, because the model earns its margin in the shipments after the first. A brand that fixes failed payments, gives customers a real way to skip and pause, and reads its cohorts properly has usually found more revenue than another round of ad spend would have produced. That work is the scope of our subscription retention service, and it starts with the ledger and the survival curve, not with a discount.
Sources
- Baremetrics: about 9% of MRR lost to failed payments (independent, across subscription businesses).
- Paddle / ProfitWell: 20 to 40% of churn is involuntary (measured across their customer bases).
- Stripe: 25% of lapsed subscriptions trace to payment failure (vendor-reported).
- The 2,000-subscriber cycle in the failed-payment section is a hypothetical for illustration and is not drawn from Pointerflow’s data or any client’s.