How to reduce subscription churn ecommerce brands accept as normal

Churn is mostly friction, not a change of heart

The story brands tell themselves is that subscribers go off the product. Sometimes they do. Far more often a card expired, a box turned up while the last one was still unopened, or somebody wanted a different flavour and found no way to say so.

That distinction decides where the work goes. A change of heart needs a better product. Friction needs an email sent at the right hour, and it's the cheapest retention work available to anybody running a subscription.

Dunning is the highest value email nobody writes

• A first attempt notice that says which card, and links straight to updating it
• Retries spaced over days, not hours, because bank declines often clear on their own
• A plain text version, because a payment email that looks like a promotion gets ignored
• A final notice that says exactly what happens next, and when
• A separate path for expired cards, which you can see coming weeks in advance

Almost every subscription platform ships a default dunning sequence, and almost nobody rewrites it. It's the one email in the account where the customer still wants to buy and the system is the only thing in the way.

The box that arrives too soon

The second cause is a cadence somebody guessed at launch and never went back to. If a customer is still working through the last delivery when the next one ships, the subscription starts to feel like a standing order they forgot to cancel.

The fix is a pre-shipment email a few days before the charge, with skip, delay and swap in the buttons. Looks like an invitation to leave. It's the opposite: it turns a cancellation into a postponement, and postponements come back. The same timing logic sits behind the replenishment flow for brands selling the same thing without a subscription.

What to offer, and in what order

Most cancel flows open with a discount, which is the most expensive option on the page and the one that teaches customers to threaten to leave. Put it last. And only for the reasons where money is genuinely the problem.

A subscription is repeat purchasing with a calendar attached

Brands with weak repeat purchasing rarely fix it by adding a subscription option. The plan makes a good habit easier and a bad one more visible, which is why the post purchase work usually has to come first. Supplement brands are the clearest case, because the reorder window is set by the pack size and not by mood.

Split the churn number before you try to move it

One churn percentage tells you almost nothing, because it adds together two problems with different owners. Involuntary churn is a payments and email problem. Voluntary churn is a product, pricing and cadence problem. Report them as one number and they cancel each other out, so nobody can tell which way anything moved.

Then read it by cohort, not by month. Everybody who started in March, tracked forward, will show you where the plan loses people: in the first delivery, at the second charge, or slowly over a year. Three different fixes, and a monthly average hides all of them.

What to build, and in what order

None of this needs a new platform. It needs five emails written properly and a cancel page that offers something before it offers a refund.

1. Rewrite the dunning sequence, because it's the only place where the customer still wants to pay
2. Add the card expiry warning, which is the one churn event you can see coming
3. Send a pre-shipment email with skip, delay and swap in it
4. Rebuild the cancel page so pause and swap come before any discount
5. Send a real onboarding sequence in the first month, because the earliest cancellations are people who never understood the plan
6. Only then build the winback for people who have already gone

Order matters, because the first four stop losses that are happening this week and the last one works on people who have already left. Most programmes get built in exactly the reverse order. The full picture sits on subscriptions and loyalty.