Every cancel click comes with a reason, and most brands never find out what it was. The subscriber picks "cancel," gets a generic "are you sure?" with a discount, and either takes the discount or leaves. A good subscription cancellation flow does something more useful. It asks one question, reads the answer, and puts the one offer in front of the subscriber that solves their actual problem.
This guide covers the full flow, from the survey inside the customer portal to the emails after someone leaves anyway. For the bigger picture on retention, start with our hub on the subscription churn email flow.
Where the cancellation flow lives
There are two halves, and they need to know about each other.
The first half sits in your subscription platform. Recharge, Loop and most other subscription apps let you set up a cancellation flow inside the customer portal, with a reason survey and offers shown before the cancel is confirmed. What exactly you can configure depends on the app and your plan, so look at your own settings before planning anything.
The second half runs in Klaviyo. Once the cancel goes through, the platform sends a cancellation event, ideally with the reason attached as a property. That event triggers your post-cancel emails and, a few weeks later, the winback.
The most common mistake is building both halves in isolation. The portal offers 20% off, then the email an hour later offers 20% off again. Or worse, 25%. Now the subscriber has learned that leaving pays.
The reason survey
Keep it to one question with a short list of answers. Five to seven options is plenty. Each answer should map to a different save offer, otherwise there's no point asking.
A list we often start with:
- I have too much product at home
- It's too expensive
- I don't like the product or flavor
- I'm going away or need a break
- I'm not seeing results
- Shipping or delivery problems
- Something else (free text)
Adjust the wording to your product. A coffee brand has different reasons than a supplement brand. And read the free text answers every month, because that is where you find the reasons you forgot to list.
Pause, skip, change date, swap or discount
These are the five tools you have. They are not interchangeable, and the cheapest one for you is rarely a discount.
| Save offer | What it does | Cost to you |
|---|---|---|
| Skip | Drops the next order, subscription continues after | One order's revenue, no margin hit |
| Pause | Stops orders for a set period, then restarts | A few orders' revenue, no margin hit |
| Change date or frequency | Moves the next order or spaces orders out | Slightly lower revenue per month, subscriber stays |
| Swap | Different product, flavor or size | Usually none |
| Discount | Lower price on the next order or a few orders | Direct margin hit |
A skip or a longer interval keeps the subscription alive at full price. A discount keeps it alive at a loss on that order. Both count as "saved" in your dashboard, but only one of them is good business.
Match the save offer to the reason
This is the part most cancel flows get wrong. The offer should solve the problem the subscriber just told you about.
| Reason | First offer | Backup offer |
|---|---|---|
| Too much product | Change frequency | Skip the next order |
| Too expensive | Smaller size or longer interval | Discount on the next order |
| Don't like it | Swap to another product or flavor | None, let them go |
| Going away | Pause until a chosen date | Skip |
| No results yet | Usage tips and realistic timing | Swap |
| Delivery problems | Fix it and say so personally | Change date |
Show one offer, maybe two. A wall of six options looks like you're trying to trap someone, and it makes the cancel harder to find. If they decline, let them cancel with one more click.
On price: before you reach for a discount, check whether a cheaper option already exists in your range. Subscribers who say "too expensive" often just bought a bigger size than they need. A smaller pack at full price beats a discounted big one.
Keep cancelling easy
Showing a save offer is fine, as long as the way to cancel stays easy to find. Many countries and US states have rules about how easy cancelling must be, and they keep changing. Have your legal team look at the flow. Apart from the law, a subscriber who has to fight to cancel won't come back, and they'll tell people about it.
The emails after someone cancels
Some people will cancel no matter what you offer. The email side starts there.
Cancellation confirmation
Send it within minutes. Confirm the cancel clearly, say whether any order is still on its way and when, and add a single line on how to restart. No discount here. The subscriber just said no to your offers, and an email that pushes again feels like you weren't listening.
Check whether your subscription platform already sends this confirmation. If it does and you add one in Klaviyo, the subscriber gets two.
One personal follow-up
A day or two later, send a short plain text email from a real person on the team. Ask what would have made them stay, and ask them to just hit reply. You will not get many answers. The ones you get are worth more than the survey, because people write in full sentences.
If the reason was a fixable problem, like a delivery issue, this email is where you say what you changed. Branch the flow on the cancellation reason property if your platform sends it.
Winback timing
Don't start the winback the day after the cancel. The subscriber probably still has product at home. Start it around the time the last order would be used up, which for most monthly products is three to six weeks after the cancel.
Then space a few emails over the next one to two months. Lead with something that has actually changed: a new flavor, a smaller size, a price change, a fixed problem. Link straight to reactivating the old subscription where your setup supports it, so the subscriber doesn't have to start over.
Use the cancel reason here too. People who left over price can get your best offer in the winback. People who left because they had too much product just need a reminder and a lower frequency. Our winback flow guide goes deeper on offers and spacing.
What to measure
- Save rate per reason. Not one overall number. If "too expensive" saves well and "don't like it" never does, you know where to put the work.
- Saved subscribers 60 days later. A save that cancels again after one more order wasn't much of a save.
- Discount share of saves. If most saves come from discounts, try the cheaper options first.
- Reactivations from the winback. Check this against the cancel reason to see which reasons are worth chasing.
Cancellation is only one source of churn. Card declines quietly cost subscription brands a lot too, which is why we always pair this flow with failed payment recovery emails. If you'd like us to set up the whole retention side for your brand, see our subscription marketing service.
Questions we get a lot
What is a subscription cancellation flow?
It is the set of steps a subscriber goes through after clicking cancel: usually a short reason survey, a save offer that fits the reason, a confirmation, and then follow-up emails. Part of it lives in the subscription platform's customer portal and part of it in your email tool.
Should every cancelling subscriber get a discount?
No. A discount only fits people who leave because of price. Someone with too much product at home needs a longer interval or a skip, and someone who doesn't like the flavor needs a swap. Handing everyone a discount trains subscribers to cancel for a better price.
When should the winback flow start after a cancellation?
Most brands wait a few weeks, roughly the time it takes for a subscriber to run out of product. Send one plain confirmation email right after the cancel, then start the winback once the last box is likely used up.
Is it legal to show save offers before someone can cancel?
In many places it is fine to show an offer, as long as cancelling stays easy and the subscriber can always get to the cancel button without a maze. Rules differ by country and state and keep changing, so have your legal team check the flow.