Cancellation Flow

Design Pattern

Make leaving easy to find and complete in a few steps, and respond to the user’s situation without blocking the exit.

Make it easy to find

Put the option where users look for it, in account or billing settings, and name it plainly. Some jurisdictions require cancelling to be about as easy as signing up.

Offer something that fits, once

Ask one optional question about the reason. Respond with something that fits the answer: a pause for low usage, a cheaper plan for price, a route to support for a technical problem, and a clean exit for everyone else. Present the offer as an option beside the exit, never in front of it.

Say what happens to the data

Explain what will be deleted and whether it can be recovered, and let users export their data before they go.

Confirm and close the loop

Ask for a deliberate confirmation (such as the account password), then show a clear confirmation of what happened. Invite feedback after the exit instead of before it, and make it optional.

Responding to the actual reason for leaving helps the users who would stay for the right offer, and respects the ones who wouldn’t. A clean exit builds trust and leaves the door open to return. It is the counterpart of a Trial Countdown at the start: both are about being honest about time and money.

Obstacles and guilt

Hidden cancel links, repeated confirmation screens, guilt-inducing copy, and offers that block the exit damage trust and can break the law. They also play on Loss Aversion, the tendency to weigh what we might lose more heavily than what we might gain, and users notice when it is used against them. A good flow lets the user leave in as few steps as they would take to join.

Account closing versus cancelling a plan

Closing an account is more final than cancelling a paid plan. The first deserves a stronger confirmation and a data export. The second can often be softened with a pause or a downgrade.

What the example does

Mailchimp’s flow starts with “Before you go…”, offers a conversation with a specialist, and offers to export the account data. It then asks for the password, closes the account, and invites optional feedback. It does not ask why the user is leaving before it offers help, so the offer can’t be tailored to the reason.

The reason question

Shopify’s flow asks the main reason for closing, with a list that covers the real ones (“struggling to get set up”, “too expensive”, “not making enough sales”). That is the information needed to answer with something that fits. Here the question is required, which turns it into a gate. Making it optional keeps the exit as easy as the pattern asks for.

Alternatives next to the exit

Netflix’s “Manage your membership” page puts “Change plan”, “Pause for up to 3 months”, and “Cancel” side by side, each with a one-line description. It states the date the change takes effect and that the user can keep watching until then, and shows what to watch before leaving. The alternatives are offered without hiding the exit. Describing the cost of leaving as it really is, here losing personalized recommendations, is framing that informs. Inventing losses to scare users is not.

A grace period

Deleting an account is permanent, so a waiting period protects users from their own mistakes. Atlassian’s confirmation after closing a Trello account names the exact date after which deletion is final and puts “Cancel deletion” at the top, as the main button. The clock is a fact about the data, and not a pressure to stay.

The way back

Shopify closes with a statement of how long data is kept and a one-step link to re-open the store. A clean exit that makes returning easy keeps the door open.

Measuring it

Track the save rate by stated reason, the return rate of former customers, and complaints about cancelling.