A subscription does not sell more, it moves the decision

The core of a subscription is simple. The customer decides once instead of every time she reorders. That removes the moment in which she forgets the product, picks up an alternative at the chemist or orders from another shop. Additional revenue only appears if that moment has actually been costing customers so far.

This is where many calculations turn out too optimistic. Shops compare the revenue of subscribers with that of all other customers and see a clear gap. The comparison is skewed. People who subscribe were usually convinced of the product beforehand. The gap therefore mostly shows who signs up, and only to a lesser extent what the subscription achieves.

A more honest yardstick is how the same customers would have behaved without a subscription. That can rarely be measured precisely. A usable approximation sits in your own order data, however: how many new subscribers had already reordered two or three times at full price before switching. The higher that share, the less additional revenue the subscription brings.

When a subscription model in an online shop raises customer value

A subscription only raises customer value, meaning the profit a customer generates over the whole relationship, under certain conditions. The more of them apply, the more likely the model is to hold up.

Even consumption. Day cream, shampoo or food supplements run out within a predictable period. For products like these, a delivery rhythm can be set that fits everyday life. For products used quickly one month and not at all the next, that does not work.

Repurchase without a wish to switch. Many customers buy the same proven skincare product for years. With decorative cosmetics, by contrast, change is part of the appeal: new shades, new seasons. Here a subscription works against what customers actually want.

Margin after discount and shipping. Every subscription delivery carries its own shipping, packaging and payment costs. With small baskets, a single shipment can absorb a large part of the contribution margin. The shop then earns less on each delivery than on a bundled order.

A real gap between purchases. A subscription does most good where customers are lost today between two purchases. This leads to an insight that runs against intuition. A shop gains most not from its most loyal customers, but from the moderately loyal ones who would like to come back but do not do so reliably.

The discount often pays for orders that would have come anyway

Most subscription offers rely on a permanent price reduction. Suppose a shop gives ten per cent off every subscription delivery. For a customer who would otherwise have drifted away, that is a good trade. For a customer who would have reordered on time anyway, the ten per cent is pure margin loss — on every delivery, for years.

Because the most loyal customers subscribe most often, the discount disproportionately hits the second group. This barely shows in revenue figures. Revenue per customer even rises; profit just does not rise to the same degree.

There are alternatives to a percentage discount. Service instead of price: free shipping from the first subscription delivery, a sample of new products or early access to limited editions. Benefits like these often cost the shop less than they are worth to the customer. Staggered benefits: a small reduction at the start that grows with the length of the relationship. This rewards staying rather than signing up. The price of this option is that it looks less tempting at sign-up.

One checkable sign: if many subscribers cancel after the first, especially discounted delivery, the discount has attracted bargain hunters rather than retained regular customers.

The delivery rhythm decides whether customers stay

Many cancellations have nothing to do with the product. A customer likes her face cream, but three unopened jars are now sitting in the bathroom cabinet. The rhythm was too fast, the stock grows, and at some point she cancels to catch up. From the shop’s point of view, this looks like dissatisfaction.

Cosmetics add a second problem. Opened products have a limited period of use, shown by the open-jar symbol with a number of months. Anyone with too much stock ends up throwing product away and sees the subscription as waste. A rhythm that is too slow creates the opposite problem: the pack is empty, the next delivery has not arrived, and the customer buys elsewhere in between.

The rhythm should therefore come from your own data, not from the default setting of the subscription software. The intervals between past reorders of a product show how long a pack actually lasts. Equally important is that customers can react without friction: postpone a delivery, skip one, change the quantity. A reminder a few days before dispatch gives them the occasion. How such emails should be written is covered in the article on transactional emails in online shops.

Freedom to cancel harms the subscription less than feared

In Germany, how easy cancellation must be is largely set by law. Since July 2022, online shops through which consumers can enter into ongoing contracts need a clearly visible cancellation button. Since March 2022, such contracts may only renew for an indefinite period after a minimum term, with a notice period of no more than one month. The exact implementation should be checked by a lawyer, as mistakes can lead to formal warnings.

Some shops still try to make cancellation harder: hidden links, multi-step confirmations, cancellation by email only. This keeps some customers subscribed in the short term. In the long term it comes back at a cost, as a payment reversal, a complaint to customer service or a public review.

The counter-calculation is often overlooked. Easy cancellation lowers the threshold at sign-up, because the risk seems small to the customer. Offering a pause function as well keeps some customers who would otherwise have cancelled. A short, optional question about the reason for cancelling is also worthwhile. It provides the data needed to improve rhythm and offer.

Failed payments end more subscriptions than expected

Some subscriptions end without anyone cancelling. The card has expired, the account lacks funds, the bank declines the charge. Many shop reports do not show this involuntary churn separately. It disappears into the overall drop-off rate, although these customers actually wanted to stay.

The causes differ by payment method. For credit cards, the first payment is confirmed with strong customer authentication under European payment services rules. Subsequent payments run as merchant-initiated transactions, yet individual banks still sometimes ask for confirmation again. With SEPA direct debit, the customer can have an authorised debit refunded within eight weeks without giving a reason. Each returned debit also incurs bank fees, and by then the goods have long been shipped.

A well-designed recovery process, known in the trade as dunning, counters this loss. It includes retries spread over time, an email with a direct link for updating payment details and a notice before a stored card expires. A clear rule on whether a delivery ships before payment succeeds matters just as much. A first step in your own shop: analyse how many ended subscriptions go back to a cancellation and how many to a payment.

When a subscription model in an online shop does not pay off

Whether a subscription will hold up can be estimated fairly well before launch. The most important clues are in your existing order data. If the intervals between two purchases of the same product vary widely, no fixed rhythm will fit. If hardly anyone buys the same product a second time, the basis is missing altogether.

Several situations argue against it. A novelty-driven range: if customers come mainly for new products, a subscription slows exactly that drive. Small baskets with a high shipping share: here every single delivery eats the margin a bundled order would still have had. No capacity for operation: a subscription needs someone to handle failed payments, queries and rhythm changes. Without that person, the model quietly decays.

The technology also deserves a sober look. Many shop systems handle subscriptions through an additional app or plugin. That brings ongoing fees and a further dependency to consider with every update. When such extensions become a burden is covered in the article on trimming Shopify apps. A sensible start is usually a pilot with one or two products that clearly meet the conditions above.

After launch, the subscription becomes an operational task

A subscription model is not a feature that is set up once and then runs. It is an ongoing contract with many customers, and its economics only become visible over months. What matters is looking at cohorts, meaning groups of customers who started in the same month. The question is how many of them are still active after three, six and twelve months, and what they have brought in after all costs.

Alongside this, failed payments, cancellation reasons and postponements need regular review. They show whether rhythm, benefits and dunning still fit. How we support shops from architecture through to day-to-day operation is described under Services.