Everyone has met the subscription that is easy to join and nearly impossible to leave. The cancel option buried three menus deep. The online sign-up that can only be cancelled by phone, during office hours. The five screens of offers, each with a small grey "no thanks" link. The final page that asks whether you are sure you want to lose all your benefits. On a dashboard, that flow can look like a retention win. In reality, it is building a queue of resentful customers.
Trapped subscribers are not retained subscribers. They dispute charges with their bank, write angry reviews, warn their friends and, increasingly, complain to regulators. The better way to think about it is this: a cancellation flow should be a conversation, not a maze. Its job is to offer the right alternative to the customers for whom one genuinely fits, and to let everyone else leave cleanly and quickly. Done that way, it saves more of the customers worth saving and protects the brand while it does.
This is general guidance for marketers, not legal advice. Subscription and automatic-renewal rules differ by country and by US state, and they have changed frequently. Check current requirements with a qualified lawyer before changing how you sell or cancel subscriptions.
The rules in 2026, briefly
Much of what is written online about cancellation law is now out of date, so it is worth being precise. In the US, the FTC finalised amendments to its Negative Option Rule in 2024, widely known as the "click-to-cancel" rule. On July 8, 2025, the Eighth Circuit Court of Appeals vacated that rule in its entirety on procedural grounds, days before its main requirements took effect. The FTC has since begun a new rulemaking process, so a replacement may follow, but the 2024 rule is not in force.
That is not the same as no rules. The federal Restore Online Shoppers' Confidence Act still requires subscriptions sold online to disclose material terms clearly, obtain express informed consent before charging, and provide a simple way to stop recurring charges, and the FTC continues to enforce it. In 2025, Amazon agreed to a $2.5 billion settlement with the FTC over how Prime memberships were sold and cancelled. Meanwhile, roughly 25 to 30 states have their own automatic-renewal laws. California's law, amended with effect from July 2025, and a Massachusetts regulation both require cancellation to be at least as easy as signing up, and several states place limits on how retention offers are presented. The practical takeaway is that the standard the vacated rule described has become the safe design target anyway.
Flippa
World's largest marketplace to buy and sell websites, online businesses, SaaS, apps and domains.
Best for: Buying & selling online businesses, sites & domains
Anatomy of a good cancellation flow
The best flows are short and follow the same basic shape, whatever the product.
- Easy to find. A clearly labelled cancel option in the account area, reachable in a couple of clicks, through the same channel the customer used to sign up.
- One reason question. A single, quick multiple-choice question about why they are leaving, with an optional comment box.
- One relevant alternative. A single offer matched to that reason, shown once, with declining made exactly as easy as accepting.
- A clear exit. A plainly worded button to finish cancelling, followed by confirmation on screen and by email, including the date the service ends.
What the flow must not contain is equally important: hidden or low-contrast cancel links, phone-only cancellation for online sign-ups, repeated offers after a customer has declined, confusing double negatives, guilt-trip wording, or extra steps whose only purpose is to wear someone down. Those tactics are exactly what regulators describe as dark patterns, and they are the same design habits we flagged in consent flows.
Match the offer to the reason
The single reason question is what makes a flow useful rather than annoying, because it lets you offer something that addresses the customer's actual situation. It is the same diagnostic logic we covered in churn diagnosis, applied in the moment.
| They say | Offer this | Avoid |
|---|---|---|
| "Not using it right now" | Pause for one to three months | A discount on something unused |
| "Too expensive" | A cheaper plan or smaller package | A steep temporary discount |
| "Too much product" | Skip a delivery or reduce frequency | Pushing the same schedule |
| "Something went wrong" | Fix it, or offer real help | Money in place of a solution |
| "Don't need it any more" | A clean, friendly exit | Any offer at all |
The pause is the most underused tool in subscription retention. Many cancellations are about timing rather than dissatisfaction: someone travelling, moving house, tightening a budget for a few months or simply overstocked. A pause solves their real problem without asking them to pay for something they don't need, and it keeps the relationship open. Downgrades work similarly for price concerns, and usually retain customers far better than a deep discount on the same plan, because the customer ends up on something that genuinely fits.
The principle A cancellation flow should be a conversation, not a maze. Offer the right alternative to the customers for whom one genuinely fits, and let everyone else leave in a couple of clicks. Trapped subscribers are not retained subscribers.
Why blanket discounts backfire
Offering a large discount to everyone who clicks cancel is the most common retention tactic, and one of the least effective over time. It addresses only customers for whom price was genuinely the problem, and even then it usually delays rather than prevents cancellation: when the discount ends, many cancel again. Worse, it teaches customers that threatening to leave is the way to get a lower price, and that lesson spreads. Customers comparing notes online quickly discover that the cancel button is a coupon.
If you do use a price-based offer, reserve it for the customers who cited cost, keep it modest, and make the terms clear, including what happens when it ends. It is the same margin discipline we applied to loyalty rewards: a discount that retains no one who would otherwise have left is pure cost.
Measure saves that last
The number most teams report, the share of cancel attempts that accept an offer, flatters almost every flow. What matters is how many of those customers are still paying later.
1,000 subscribers start the cancellation flow in a month.
Flow A: 50% discount for everyone
→ 35% accept the offer = 350 "saves"
→ Only 40% are still paying 90 days later → 140 real saves, all at half price
Flow B: offers matched to reason (pause, downgrade, skip)
→ 25% accept an offer = 250 "saves"
→ 70% are still active 90 days later → 175 real saves, mostly at full or right-sized price
Flow A looks better on the day. Flow B keeps more customers, at better margins, with fewer disputes. Report the 60- or 90-day net save rate, split by reason and offer, not the click-through on the offer screen.
Watch the side effects too. Chargebacks, payment disputes, complaints and reviews that mention cancellation are the early warning signs of a flow that has tipped from helpful into obstructive. A rising immediate save rate alongside rising disputes is not success; it is a problem arriving later. Tracking these outcomes by cohort, as described in cohort analysis, shows whether saved customers really behave like retained ones.
Before and after the button
The cancellation flow is the last line of defence, not the first. The reasons customers give there are some of the most useful feedback you will get, so store them properly as part of your first-party data and feed them back into onboarding, product and pricing so that fewer people reach the button at all. Clear terms at sign-up, stated plainly in the same spirit as a readable privacy policy, reduce the number of people who feel surprised later.
Before renewal, send clear reminders, especially for annual plans and trials converting to paid, by email or SMS. It sounds counterproductive, but surprise renewals are a major source of disputes and resentment. After a customer leaves, end on good terms: a friendly confirmation, no guilt, and an easy way back. Customers who leave well are far more likely to return, and a thoughtful sequence later, segmented by the reason they left, is exactly the job of re-engagement automation and good email segmentation. The retention maths explains why this matters so much, as we showed with recurring revenue: every month a customer stays compounds. Review the whole flow regularly as part of a funnel audit.
The short version
A cancellation flow should be a conversation, not a maze. Trapped subscribers dispute charges, leave angry reviews and draw regulatory attention, so obstruction is a false economy. On the rules: the FTC's 2024 click-to-cancel rule was vacated by the Eighth Circuit in July 2025 and a new rulemaking has begun, but ROSCA still requires clear terms, informed consent and a simple way to stop charges, enforcement continues, and roughly 25 to 30 states have their own automatic-renewal laws, several requiring cancellation to be as easy as sign-up. Design the flow to be easy to find, ask one reason question, show one relevant alternative that is as easy to decline as to accept, and end with clear confirmation. Match offers to reasons: pauses for timing, downgrades for price, skips for oversupply, fixes for problems and a clean exit when the need has ended. Avoid blanket discounts, which delay rather than prevent cancellation and teach customers that the cancel button is a coupon. Measure the 60- or 90-day net save rate by reason, watch disputes and complaints, send renewal reminders, and let customers leave on good terms.
Want a cancellation flow that saves customers fairly?
We help subscription brands design retention flows customers trust.
Explore E-commerce Marketing →Frequently asked questions
Is the FTC click-to-cancel rule in effect in 2026?
No. The FTC finalised amendments to its Negative Option Rule, widely known as the click-to-cancel rule, in 2024, but the US Court of Appeals for the Eighth Circuit vacated the rule in its entirety on July 8, 2025, days before its main requirements were due to take effect, on procedural grounds. The FTC has since begun a new rulemaking process, so a replacement rule may follow, but it was not in force as of late 2026. That does not mean cancellation practices are unregulated. The federal Restore Online Shoppers' Confidence Act still requires clear disclosure of material terms, express informed consent before charging, and a simple mechanism to stop recurring charges for subscriptions sold online, and the FTC continues to bring enforcement actions under it. Roughly 25 to 30 states also have their own automatic renewal laws, and several, including California and Massachusetts, now require cancellation to be at least as easy as signing up. This is general information, not legal advice.
What makes a good subscription cancellation flow?
A good cancellation flow is easy to find, short, honest and useful. The cancel option should be clearly visible in the account area, available through the same channel the customer used to sign up. The flow should ask one simple question about why the customer is leaving, then present a single relevant alternative based on that answer, such as a pause for someone who is not using the product right now or a cheaper plan for someone concerned about cost. Declining that offer should be as easy as accepting it, with a clear button to continue cancelling, and the process should end with an unambiguous confirmation on screen and by email. What it should not include is hidden buttons, phone-only cancellation for an online sign-up, repeated offers, confusing wording, guilt-trip messaging or extra steps designed purely to wear the customer down. The aim is to help the customers for whom an alternative genuinely fits, and let everyone else leave cleanly.
What is the best save offer for a cancelling subscriber?
The best offer is the one that matches the reason the customer gave, which is why asking a single reason question first matters so much. For customers who are not using the product at the moment, a pause of one to three months is often the most effective option, because it solves their actual problem without asking them to pay for something they do not need. For customers concerned about cost, a cheaper plan or a smaller, less frequent delivery usually works better than a deep discount on the same plan. For customers who have too much product, skipping a shipment or reducing frequency fits. For customers reporting a problem, the right response is to fix it or offer help, not money. For customers whose need has genuinely ended, there is usually no suitable offer, and the respectful move is to let them go. Blanket discounts offered to everyone tend to delay cancellation rather than prevent it, and they can teach customers that clicking cancel earns a price cut.
How do you measure whether a cancellation flow works?
Measure customers who are still paying some time after the save, not just customers who clicked an offer. Many save offers look impressive at the moment of acceptance and then quietly fail when the customer cancels again once the discount ends or the pause runs out. A more honest measure is the net save rate: the share of people who started cancelling and were still active and paying, typically 60 or 90 days later. Track it separately for each reason and each offer, because a pause might retain one group well while a discount only delays another. Also watch the side effects: chargebacks and payment disputes, complaints, negative reviews that mention cancellation, and how many customers attempt to cancel repeatedly. A flow that produces a high immediate save rate alongside rising disputes and complaints is not working; it is storing up problems. The reasons customers give should also feed your wider churn analysis so that fewer people reach the cancel button in the first place.
