Skip to content

Moves: Repeat purchases

Subscription & retention marketing

Turn one order into a schedule: the right subscription model, easy pausing and swapping, failed payments recovered, retention read by cohort.

Get your free growth plan
Subscription & retention marketing: subscribe Written on Cardboard Being Held by Group of People.

Who it's for

  • Brands selling something that runs out on a predictable cycle
  • Stores already offering a subscription that few customers choose
  • Merchants with high cancellation in the first two or three cycles
  • Businesses whose repeat rate is good but unmanaged and unmeasured
  • Sellers of durable goods looking for a second purchase that makes sense

Retention is the lever that compounds. A customer who comes back was not paid for twice, so every point of repeat rate is worth more than the same point anywhere else in the business. Subscriptions are the strongest version of that, because they replace a decision with a schedule.

They are also easy to get wrong. A subscription put on a product nobody rebuys creates cancellations and support tickets. A programme with a saving too small to notice gets ignored, and one with a saving too large gives away the margin the repeat order was supposed to earn.

So we start with what your order history shows people actually buy again, choose a model that suits it, and design the management experience as carefully as the sign-up. Then we deal with the churn nobody talks about, which is payments failing quietly on expired cards.

For stores whose products are not consumable, the same lever is still there. It just runs on replenishment cycles, accessories, upgrades and reasons to return rather than on a recurring charge.

Repeat purchases is the lever all of this pulls on. The free plan scopes what we would build and what the base fee plus growth share would be to run it.

How we run it

The honest process, in order.

  1. Work out what people genuinely rebuy

    Order history answers this before any strategy does. We look at which products appear in a second order, how long the gap is, which first purchases predict a second one, and which products are bought once and never again for perfectly good reasons. A subscription belongs on the first group only. Putting one on the wrong product creates cancellations, refunds and support work, and it teaches your team that subscriptions do not work when the offer was simply attached to the wrong thing.

  2. Choose a model that suits how the product is used

    Subscribe and save fits a staple someone already buys on a rhythm. A curated box fits discovery, where the appeal is the surprise rather than the supply. A prepaid term fits products where the benefit only appears after several months, and it trades a discount for commitment. Each has different economics and different churn behaviour. We pick one to start with rather than launching three, and we set the saving from your margin rather than from what looked normal elsewhere.

  3. Make managing it as easy as starting it

    Every subscriber should be able to skip a delivery, change the date, swap the product, change the quantity and cancel, without emailing anyone. That sounds like it invites cancellation and it does the opposite, because most people who cancel are really trying to solve a smaller problem: too much stock, wrong size, bad timing. Give them the smaller fix and they stay. Hiding the cancel button produces chargebacks, angry reviews and a payment processor asking questions.

  4. Recover failed payments before chasing cancellations

    A large share of subscription losses are not decisions at all. A card expires, a bank declines a routine retry, and the customer never knows the order stopped. We set a retry schedule spread across days rather than hammering the same card, enable the network updater so replaced cards carry over, and send a message that makes updating the card take one tap. This is the cheapest retention work available and it is usually the least attended to.

  5. Give the second and third order their own reason

    The gap between the first and second purchase is where most customers are lost, and a generic newsletter does not close it. We build a sequence tied to the product bought: how to use it well, what to expect in the first weeks, what goes with it, and when to reorder. For durable goods the second purchase is a companion item or a consumable part, so the flow points there rather than at the same product again.

  6. Read retention by cohort, not by month

    A monthly retention percentage moves when acquisition moves, which makes it useless for judging whether anything improved. We group customers by the month they first bought and follow each group over time, so you can see whether people who joined after a change stayed longer than people who joined before it. Cohorts take months to say anything, which is exactly why they should be started now rather than when someone asks for them.

Worked example, not a result

Example: what a fifth cycle is worth

Say a store signs up 100 new subscribers in a month at $40 a delivery. If 60 are still active at the second delivery, 45 at the third and 35 at the fourth, that cohort is worth $9,600 over four cycles.

Now say failed payments account for 10 of the losses across those cycles, and card updating and better retries recover 6 of them. Those 6 people carry on at $40, and the same fix applies to every cohort after this one.

The extra revenue is small in month one and compounds, because each recovered subscriber keeps paying while new cohorts stack on top. Round made-up numbers, used to show why cycle-by-cycle retention matters more than the sign-up count.

Questions about subscription & retention marketing

Some of it, yes, and that is usually a good trade. A customer who was already buying every eight weeks moves onto a schedule, so you give up a small discount and gain predictability plus a much lower chance of them drifting to someone else. It stops being a good trade when the saving is large enough that loyal customers move onto it and buy no more often than before, which is why the discount comes out of the numbers rather than out of the air.

There is no honest benchmark to give you, because it depends on the product, the price, the cycle length and how you acquired the customer. What matters is the shape: most cancellations cluster in the first few cycles, and the curve flattens after that. We measure your own curve, work on the earliest drop first, and compare each cohort against the ones before it.

Usually not first. Points programmes cost money to run, take a long time to influence behaviour, and often reward customers who were coming back anyway. Flows, replenishment and subscriptions do more for less. If your category genuinely competes on loyalty and your repeat rate is already healthy, it becomes worth modelling, and the plan will say so rather than assuming.

Yes, though not as a recurring charge. Durable goods retain through accessories, replacement parts, care products, upgrades, and the next item in the range. The cycle is longer and less predictable, so the work moves toward knowing when a customer is likely to be ready and having something worth telling them at that point.

Whichever one your platform supports well and your team can actually operate. What we care about is whether it lets customers manage their own schedule, handles failed payments sensibly, and reports by cohort. If your current tool does those three things, we use it. We only recommend changing when the tool itself is the reason customers are leaving.

Niches where this matters most

See how subscription & retention marketing fits your store.

The plan scopes this service for your store before anything starts.

Store, name, email and phone. A person replies from a real inbox. Privacy

Get your free growth plan