Should your brand offer subscriptions?

Subscriptions work when customers already buy again and again on a predictable cycle. They fail when a brand adds them hoping to create that habit. If your repeat purchase data doesn't already show a rhythm, a subscription program won't create one. It'll just give a discount to customers you were going to keep anyway.

This is the check we run before taking on subscription work, because roughly a third of the brands that ask for it shouldn't build it.

What predicts success?

One thing matters more than anything else: does the product get used up on a schedule the customer doesn't control?

Think supplements, coffee, pet food, skincare, razor blades and household consumables. The customer runs out whether they were thinking about you or not, and the subscription takes a chore off their plate.

Compare that with apparel, home goods, accessories, or anything someone buys because they wanted it that day. Nothing runs out, so there's no natural cycle, and a subscription asks people to commit to a decision they'd rather make fresh each time.

Category churn figures line up with this exactly. Replenishment categories run at around 5–8% a month, while curation boxes run at 10–15%. The gap comes down to whether the product runs out, not how well it's executed.

How do I check before building anything?

Your existing order data can answer three questions right now.

What share of customers have ordered the same product more than twice? If it's under about 20%, you don't have a replenishment product, whatever your category might suggest.

How tightly clustered is the gap between repeat orders? Take your repeat buyers and look at how the days between orders are spread out. A tight cluster means there's a real cycle you can build a delivery schedule around. A flat spread means people reorder whenever they happen to think of you, which is a different problem, and one email solves better than subscriptions do.

What's the margin on the products people buy again? Because that's what you're about to discount.

What does it cost?

More than most people budget for, and the platform fee is the smallest part of it.

The subscription discount is permanent, and it goes to your most loyal customers, who would have paid full price anyway. At a typical 10–15%, that comes straight off the contribution margin of your best customers.

Then there's the operational side. Failed payments need dunning. Subscribers want to pause, skip, swap and change dates, and if they can't do it themselves, they email support. Fulfillment takes on a recurring commitment that has to hold up during your busiest week. Someone has to own all of that, and it never stops.

Before you start, run the numbers: subscriber lifetime value at a realistic churn rate and your discounted margin, compared with a one-time customer's value at full margin. If that doesn't come out clearly positive with a churn assumption you'd be willing to defend, the program is really just a discount scheme with extra steps.

When is the answer clearly no?

There are four situations where we'd advise against it.

• Considered, occasional purchases. Furniture, electronics, anything where the customer wants to choose each time.
• Thin margins. Under about 40% gross, the discount plus the operational cost usually wipes out the benefit.
• No repeat behavior today. Subscriptions amplify a pattern that's already there. They can't create one.
• Nobody owns retention. A subscription program without an owner slowly falls apart. Failed payments pile up, save flows go stale, and you only find out from the churn number a quarter later.

Is there a middle option?

Yes, and not enough brands use it. Before committing to a full program, test the demand directly.

Offer a repeat-purchase reminder instead of a subscription. That's an email sent at the point your data says people run out, with a one-click reorder. It costs almost nothing, doesn't need a platform, and tests the same underlying question: do these customers want to buy again on a cycle?

If the reminders convert well, you've got evidence for a subscription. If they don't, you've saved yourself the build and found out for the price of a single flow.

What if the answer is yes?

Then build the save flow before you build acquisition. Most brands launch subscriptions with a good sign-up experience and a cancel button, then realize four months later that canceling is the only option a subscriber has ever been given.

Pause, skip, swap and change-date come first. They're what turn a cancellation into a delayed order, and they're the difference between the good end of your category's churn range and the bad end. If you're choosing a platform, our Recharge vs Skio comparison covers what changed this year.

We qualify subscription programs before we build them, and we've talked brands out of it. That's a much cheaper conversation than the rebuild. When the product fits, it can move fast: Underbrush went from no subscriptions to a major channel within a year.