Chat to us, before you go
When cancelling your subscription to Emma, it’ll offer to connect you to human support staff, encouraging you to speak to them instead of churning.
Design Examples
Cancellation Flow Examples: How Apps Reduce Churn
See how companies like Canva, Strava and Uber reduce churn, with our annotated cancellation flows featuring Loss Aversion, Framing and tailored retention offers.
Real world examples:
Updated: 26/10/2026, view all 1,508 examples
Examples on Built for Mars include
Canva lists features your team would lose
Strava counters price objections with annual
Emma shows real support staff before you go
An anti-churn tactic
If you go to cancel Uber One, they'll remind you how much money you've saved so far that month.
Tackling the price objection
If you're cancelling your Strava subscription because of price, they'll dynamically remind you that you can save money by switching to the annual plan.
Reframing your goals to reduce churn
If you go to cancel your subscription, Strava will use both Loss Aversion and Framing to remind you about your goals.
Contextual app events
As an anti-churn measure, Opal has a contextual "unblock all apps", with a motivational message, if you go to delete their app.
Creating loss aversion
Subscribe to unlock this example
Plans start from £8 per month
I don't want to forget
Subscribe to unlock this example
Plans start from £8 per month
There are more than 1400 UX Bites like this
Subscribe to BFM+
If features can be coded in minutes, thoughtful design is what wins
- Unlock the entire library
- Create custom UX courses
- Access our AI tools
£8/monthbilled annually
or, wait for the free stuff
Join the free newsletter
Get the free content when it's released, straight to your inbox, no account required
"An exceptional resource for learning how to build habit-forming products that people love."
"I’d never met anybody who looks at usability and UX the way I do until I met Peter. Only he’s smarter."
"The completest view of the psychological underpinnings of human behaviour that I’ve ever seen."
"What a nice idea! Christopher Alexander would be proud to see patterns collected like this."
FAQs
How can loss aversion work in a cancellation flow without feeling predatory?
It lands best as clear, personal context at the moment of decision, rather than a generic warning. For example, Canva lists the features a team has actually used, with usage counts, and notes that a legacy price would rise if they leave. It reads as useful information, not pressure.
Do generic benefits or personal data make better retention messages?
Personal data, because the specificity of the framing is what persuades. Strava's "Don't leave your goals behind" works, but showing the user's actual goal (e.g., running 10k) would be stronger. Uber One's "You saved £2.67 this month" follows the same logic, though it only persuades when the saving is near the monthly fee.
Are cancellation reason surveys reliable enough to trigger retention offers?
They're useful but noisy, as many people enter rubbish answers just to get through. Strava's annual-plan offer for anyone citing price shows the upside. The open question is lifetime value, so it's worth testing a one-time discount against a plan switch.
Can trial-end reminders actually reduce cancellations?
Yes, because many people cancel straight away out of fear of forgetting. Komoot reassures users twice that they'll be notified before the trial ends. That removes the anxiety and uses Default Bias, so doing nothing and letting the plan renew becomes the natural choice.
How do you stop users deleting an app when they only want to change a setting?
Anticipate the reason they want to leave and offer a shortcut that fixes it. For instance, Opal, an app blocker, adds long-press quick actions like "Unblock all apps now" and "Wait! Don't delete me!". This catches people who would otherwise delete the app rather than edit every blocking rule.
Does the same framing work outside subscription cancellations?
It does, as any cancel action is a churn moment. For example, Uber's "Cancel ride?" prompt anchors users to a driver "within 10 minutes" and warns they'll have to start their request again. Pairing a reasonable anchor with a loss is far more persuasive than either element alone.