What is being measured about you
How much of this actually worked on people, and how would anyone know?
After you buy, the seller measures whether you stay, and its messages follow those numbers. This chapter explains retention metrics from the receiving end, shows how to read a cancel flow as a retention funnel and a win-back offer as a new introductory offer, and helps you set and check your own measure of whether the purchase worked.
All stages
What the seller measures
After you buy, the campaign does not end; it changes its question. Before, the seller measured whether you clicked and converted. Now it measures whether you stay: which emails you open, how often you use the product, when you skip or pause, how much you spend each month, and how likely you are to leave. The headline numbers inside the business are retention, churn, and lifetime value, the total a customer is expected to pay before leaving. You are a row in that calculation, and the messages you receive are chosen partly by where your row sits.
Notice what these numbers do not measure. They record what you did, not whether you were glad you did it. A customer who stays because cancelling is tiresome counts the same as one who stays because the product is good. That gap is why a business steering by retention alone can drift toward making leaving hard, and why your own assessment has to come from you.
The cancel flow is a retention funnel
Nowhere is the measurement more visible than when you try to cancel. A cancellation flow is designed and tested like a checkout, with the opposite goal: each screen exists to save a share of the people who reach it. Typical steps are a reminder of what you will lose (points, saved preferences, a streak, a so-called locked-in price), a survey whose answer selects the counter-offer, an offer to pause instead, a discount, and finally a confirmation button that is less prominent than the one marked keep my plan.
Some of this is legitimate. Asking why you are leaving is fair, and a pause option suits people who really are going away for a month. The line is crossed when exit is materially harder than entry: cancellation only by phone or chat when sign-up was online, repeated screens with no visible way forward, guilt-laden wording, or a flow that ends without clearly confirming that you have cancelled. The FTC staff report on dark patterns, and the literature on sludge, treat this asymmetry as a central problem. Go in with the decision already made, decline each screen, and do not stop until you have a confirmation message, which you should save.
Win-back offers
Leaving, or starting to leave, often produces the best price you will ever be offered. Retention discounts in the cancel flow, and win-back emails in the weeks afterwards, are priced from what the business expects to lose without you. Two things follow. First, the regular price you were paying was not the lowest the seller could accept. Second, a win-back offer is a new introductory offer, and everything in the first chapter applies again: it is temporary, it renews at the regular price, and its purpose is to restart the habit. Accept it only if the product was worth the regular price to you, or if you will set a new end date and keep to it.
Measure it yourself
Answer the metrics of the seller with your own. Decide what success would look like before the trial ends: how often you used the thing, what it replaced, what it cost per use, and whether you would sign up again today at the regular price. Check those numbers on the review date you set. If the answer is yes, stay without embarrassment; a retention offer does not make a good product bad. If the answer is no, cancel, whatever is on the screen.
Techniques you will meet here, and the tell
Joining took a couple of screens; leaving is routed through many, or through a different and slower channel.
The tell: You signed up online in two minutes, and cancelling requires a phone call, a chat queue, or a link you cannot find.
Each added step in the cancel flow (survey, offer, confirmation) is there because a share of people give up at it.
The tell: You must answer or dismiss several screens before a cancel button appears, and the way forward is the least visible thing on each.
The flow lists what you will lose by leaving (points, price, saved settings), because losses weigh more heavily than equal gains.
The tell: The screen is headed you will lose, and the items listed have no value to you once you have left.
Money already spent and loyalty points already earned are presented as reasons to keep paying.
The tell: You catch yourself thinking that you have put too much in to stop now, though none of it comes back either way.
Points, tiers, and locked-in prices raise the felt cost of leaving. They are fair when the rewards are real and usable.
The tell: The points can only be spent by staying, or the locked-in price is simply the regular price under another name.
Copy and images suggest that leaving hurts someone: the team, a farmer, your own good intentions.
The tell: The flow says we will be sad to see you go, or the button to leave is worded as giving up on a goal.
Continuing is the default and requires nothing; a pause offer keeps the subscription alive and resumes billing by itself.
The tell: Pause is the large button, cancel is the small link, and the pause ends automatically with a new charge.
Saved recipes, history, and personalisation are framed as possessions you would be giving up.
The tell: You are warned that your data and preferences will be deleted, as a reason to keep paying for the service.
Worked example
Dana and Hearthbox are fictional. Three boxes have arrived, and it is the review date Dana set. She paid 173.85 dollars for 18 servings. Before signing up she wrote down her own measure: servings actually eaten, takeaway nights replaced, and cost per serving eaten at the regular price.
- 1She tallies. They cooked and ate 14 of the 18 servings; two recipes went unused in a busy week. Takeaway fell from about three nights a week to one. At trial prices each serving eaten cost about 12.42 dollars. At the regular price the same three weeks would have cost 245.79 dollars, about 17.56 dollars per serving eaten, which is more than the 16 dollars she pays for takeaway.
- 2Her answer at the regular price is no, so she opens account settings and chooses cancel, before the cutoff for the fourth box. The first screen says she will lose 400 Hearth Points and her locked-in price. The points cannot be used once she leaves, and the locked-in price is the regular price.
- 3The second screen is a survey; she selects too expensive. The third offers a four-week pause on a large button, with continue cancelling as a small link beneath. She takes the link.
- 4The fourth screen offers 30% off the next two boxes: 50.36 dollars plus shipping, 60.35 dollars each, returning to 81.93 dollars afterwards. She notes that this price was never offered while she was simply paying, and declines.
- 5On the last screen keep my plan is the bright button and cancel subscription is grey. She presses the grey one, sees a message that the subscription is cancelled and no further boxes will be sent, takes a screenshot, and saves the confirmation email.
- 6Over the next two weeks three emails arrive offering 50% off her next box. She recognises a new introductory offer and leaves it.
Outcome. By her own measure the kit worked as a three-week trial and fails as a weekly habit at the regular price. Leaving took five screens and three refusals, but the flow did let her finish online and gave a clear confirmation, which she kept. She also keeps the recipe cards.
Measurement worksheet: did it work for you?
Fill in the first four fields on your review date. Fill in the rest if you go on to cancel. Saved in this browser only.
Write what would make this worth the regular price: how often used, what it replaces, a cost per use you accept.
Record how much you used in the period, from a count and not from memory if you can.
Divide what the same period would cost at the regular price by the number of times you actually used it.
Answer for the regular price, as if you were not already a customer.
Tick every step you met between choosing cancel and receiving a confirmation.
Write the price offered, how long it lasts, and what it renews to afterwards.
Record what proof of cancellation you hold.
Decision check: staying by choice, or by friction?
0/6If any of these is not true: If your own measure says no, cancel now, before the next billing cutoff, declining every screen until you reach a confirmation, and save it. If the flow blocks you, send a written cancellation to the support address of the seller with the date and your account details, keep a copy, and go on to the steps in the final chapter. If a retention offer tempts you, write down its end date and renewal price and set a new reminder before you accept.
Common mistakes
- ×Stopping at the pause screen and believing you have cancelled. A pause ends by itself, and billing resumes.
- ×Counting money already spent, or points already earned, as a reason to continue. Neither comes back whether you stay or go.
- ×Treating a win-back price as a reward. It is a new introductory offer with the same renewal behind it.
- ×Leaving without proof. Without a saved confirmation, a later charge becomes your word against a system log.
- ×Over-suspicion. One question about why you are leaving, or a real pause option, is not a trap, and staying with a product you use and value at its regular price is a sound decision, not a manipulated one.
Did the change you wanted happen, and what did it cost the audience?
Related
References
- Sunstein, C. R. (2021). Sludge: What Stops Us from Getting Things Done and What to Do about ItDescribes how added friction, such as difficult cancellation, keeps people in arrangements they would otherwise leave.
- Federal Trade Commission (2022). Bringing Dark Patterns to Light. FTC Staff ReportIdentifies cancellation paths that are harder than sign-up, and repeated save attempts, as practices that trap consumers in subscriptions.
- Brignull, H. (2023). Deceptive Patterns: Exposing the Tricks Tech Companies Use to Control YouDescribes hard-to-cancel designs and obstruction in subscription services.
- Kahneman, D., & Tversky, A. (1979). Prospect Theory: An Analysis of Decision under Risk. EconometricaEstablishes that losses weigh more heavily than equivalent gains, the basis of you-will-lose framing in cancel flows.
- Arkes, H. R., & Blumer, C. (1985). The psychology of sunk cost. Organizational Behavior and Human Decision ProcessesShows that people continue an endeavour because of money already spent, even when continuing is not worthwhile.
- Samuelson, W., & Zeckhauser, R. (1988). Status quo bias in decision making. Journal of Risk and UncertaintyShows that people disproportionately stick with the existing arrangement, which favours any subscription that continues by default.