Subscription inertia
A Socratic walk-through of subscription inertia — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why do people keep paying every month for a service they know they no longer use?
Here is the odd part of this, which the usual explanation misses. The person paying for the unused gym membership is not confused. Ask them and they will tell you the exact amount, that they have not been since March, and that they ought to cancel. They know. They pay anyway, for years.
That rules out ignorance, which means calling it laziness is doing no explanatory work either — it just renames the thing. So what is the actual structure of the decision such that a fully informed person keeps making it the same way?
Reasoning it through
REASONING #Begin with a question that sounds pedantic and is not: when, exactly, does the decision get made?
For a purchase, there is a moment. You stand at the till and either hand over the money or you don't. A subscription has no such moment. The card is debited by standing instruction, nothing arrives asking whether to continue, and the transaction completes without any act on your part. Continuing is the default; stopping requires an action. So the decision is not made monthly at all — it is made only on the occasions when you happen to raise it yourself, and those are rare.
Now suppose you do raise it. What are you actually comparing? Not, as the naive account assumes, the annual total against nothing. You are comparing this month's nine pounds against the twenty minutes of unpleasantness needed to stop it right now — finding the account, hunting the buried link, sitting in a phone queue, declining two retention offers. And for a person with a full evening, twenty minutes of that can honestly be worth more than nine pounds.
That is the mechanism, and it is a threshold effect rather than a bias. The charge has been sliced into pieces small enough to fall below the level at which anyone mounts an effort, while the effort required to stop it has been left in one lump. Each individual month's decision is defensible. The sum of them is absurd. Notice the prediction this makes: an annual charge, arriving as one large debit, gets cancelled far more often than the same money split into twelve — because it lands above the threshold.
Two more things sit on top. The friction is not accidental. Signing up takes two clicks and cancelling takes a phone call during business hours, and that asymmetry is designed, not inherited — it has a name in the literature, sludge, and regulators in several countries have written rules requiring exit to be as easy as entry, with the American attempt finalised in 2024 and then struck down on procedural grounds. And there is optimism: the classic study of American gym members found that people on monthly contracts visited so rarely that they paid considerably more per visit than the pay-as-you-go price available at the same gyms, which they could see. They were not buying access. They were buying the belief that next month would be different.
The analogy
THE ANALOGY #Think of a tap dripping in a house you own. Each drip is worth nothing, and fixing it costs a Saturday. So it drips for four years, and on no individual Saturday was the wrong choice made.
The drip is visible and audible every time you pass it, whereas a card debit is silent and arrives on a statement nobody reads — and crucially, no one has arranged your plumbing so that the tap is deliberately harder to fix than it was to install.
Clarifying the model
THE MODEL #Three refinements connect those steps.
The first is that inattention here is not a failure of attention. Attention is a scarce resource being allocated sensibly: a nine-pound recurring charge does not merit a monthly review, and the person who audits every standing order is losing more in time than they recover. The trap is that the sensible policy of ignoring small recurring charges is exactly the policy a subscription is priced against.
The second is that the firm is not a bystander. Retention economics are built on the expected number of months after usage stops, so the fee, the trial length, the debit date and the cancellation path are all chosen with that number in mind. This is why "we'll remind you before we charge you" is a competitive differentiator rather than a courtesy.
The third is a genuine caveat: not all of this is a mistake. Keeping a service you rarely use can be rational insurance against needing it suddenly, and re-joining often costs more than never leaving. Some economists therefore read the whole pattern as a rational-inattention equilibrium rather than an error, and how much of it is really waste is contested. The test that separates the two is simple and personal: if a cancellation took one click, would you still be paying?
A picture of it
THE PICTURE #How to readRead left to right through the life of one subscription; the height of each step is how pleasant it is to go through. The shape is the argument — the entrance is the highest ground on the chart and the exit is the lowest, and that gradient is engineered rather than natural. The middle section is where the threshold does its work: noticing the charge scores low not because it hurts but because it is barely worth registering, so it never triggers the climb out on the right.
What became clearer
WHAT CLEARED #Nobody is choosing to pay for something they do not use. They are declining, month after month, to pay a one-off cost of effort in order to stop a charge deliberately sized to sit below the level at which effort feels warranted. Continuing is what happens when nothing happens, and the price of stopping is paid in a currency — a free evening — that no one is tracking.
Where to go next
ONWARD #- Why an annual renewal notice changes cancellation rates so much more than a monthly one.
- How the same slicing works in reverse, when firms split a large charge to keep it under a threshold of scrutiny.
Key terms
TERMS #| Term | What it means |
|---|---|
| Default option | the outcome that occurs when the person takes no action, which decides most recurring choices simply by being the path of no effort. |
| Sludge | friction deliberately added to a process to discourage people from completing it, the mirror image of a nudge. |
| Rational inattention | the idea that ignoring small decisions is itself an efficient use of limited attention, which is why the pattern is not simply irrational. |
Every term the collection defines is gathered in the glossary.