Your $12 Subscription Is Really $6 Every Time You Open It

There’s a question I ask about every recurring charge, and it’s so simple it feels almost embarrassing to write down: how many times did I actually use this in the last month, and what did that make each use cost? The reason the question matters is that a subscription is emotionally priced as a flat “only $12 a month”, but it is actually priced per use, and the two numbers tell completely different stories. A streaming service you opened twice costs six dollars a watch. A meal-planning app you forgot you had costs a small fortune per meal, because you had zero meals. The flat price hides the per-use price, and the per-use price is the only one that tells you the truth about whether a subscription is paying for itself or quietly robbing you.

The first time I did this properly I mostly felt silly, because the calculation is arithmetic any child could do and I had somehow never done it. You take the monthly price, and you divide it by the number of times you used the thing. That’s the entire method. What made it land was doing it honestly — not assuming I’d used something “regularly”, but actually counting, the way the recurring-charge inventory forces you to actually list what you’re paying rather than approximate from memory. The gap between how much I thought I used things and how much I actually used them was the whole story.

Here’s the mechanism, because the numbers themselves are less important than what they reveal. A service that costs $12 a month and gets used twice has a cost-per-use of $6. The same $12 used twenty times is sixty cents. Either might be worth it — a $6 gym-adjacent class you actually loved is a bargain, sixty cents for a scroll you barely remember is a theft. The point isn’t that expensive-per-use is always bad. It’s that you cannot make that judgement at all while you’re still paying the flat rate, because the flat rate is designed to feel harmless. Small subscriptions feel affordable precisely because they’re small, and the per-use maths is what breaks that spell.

Here’s the actual process, because “do the division” is too vague to survive a busy month. Once a quarter, pull the last three months of your card statements and copy out every recurring charge onto a single list. For each one, count how many times you actually used it over the most recent thirty days — not “roughly”, but by looking at your own history where you can. Divide the monthly price by that count. Then apply one rule to the result: if the per-use price is higher than what you’d pay to buy the thing one-off when you really needed it, cancel and rebuy later if you have to. That one rule is the whole of the method, and it takes about an hour a quarter, most of which is just the facing of the statements.

There’s a specific category where this calculation is brutal and worth being especially honest about, which is the subscription you keep in case you need it. Cloud storage you’re over, an editing tool you used for one project, a premium feature you meant to learn. The “just in case” subscription almost always has a cost-per-use of infinity — you’ve used it zero times — and yet it’s the hardest to cancel because the possibility of needing it feels like value. Framing it as “what does it cost per time I actually used it this month” turns that vague possibility into an answer that is often just “division by zero”.

The honest objection is that not everything is used evenly, and a per-use calculation punishes the thing you use in bursts. A tax-prep tool you use once a year, a travel app you use hard for one month, a backup service that runs in the background — none of these get a fair hearing from a single month’s arithmetic. So the method needs a second beat: instead of one month, ask over a year, and instead of raw use, ask what the year’s total use bought you. The annual framing actually rescues some subscriptions that a single bad month would wrongly kill, and it’s a useful guard against the over-correction of cancelling in a sulk. What it doesn’t change is the ones that show up as zero over twelve months, which have no defence left.

There’s one category where this method is actively misleading, and I want to flag it rather than let you discover it the hard way. Some subscriptions aren’t for using; they’re for not needing to use them. Insurance, a backup, a security product — the value of these is the absence of a bad event, so “I used it zero times” is the good outcome, not the indictment. Applying cost-per-use to them produces division by zero and the wrong conclusion every time. Save the calculation for consumption subscriptions — the things you open and use — and judge the protection ones by an entirely different question, because the method measures use, and some things you buy precisely so you never have to use them.

This is also where the per-use lens collides with the annual plan, and the collision is usually the moment of truth. An annual plan lowers the sticker price per month but it commits you before you know your real usage, which means you can end up pre-paying a year of a thing you use for three months. The discount is only a saving if the per-use price, across a real year, comes out lower — and the seller is counting on you never doing that division. Annual discounts are a loan against usage you may never spend, and the per-use calculation is how you check whether the loan ever pays itself back.

The part people find hardest, and the part that actually moves money, is converting the answer into an action. It’s one thing to know a subscription costs $15 per use; it’s another to walk over and end it, and the gap between knowing and acting is where subscriptions live forever. What finally worked for me was a rule with no room to negotiate: if the cost-per-use is above the price of simply buying the thing one-off when I actually needed it, I cancel and rebuy later if I have to. The pain of the cancellation flow is real and it’s designed to stop you, but the arithmetic gives you something to hold onto while you click through it: the subscription isn’t costing $12, it’s costing $15 a use, and you’d never have paid that at the till.

There’s a subtler version of this that I think is actually more useful than the headline calculation, which is tracking the trend rather than the snapshot. A subscription that costs $2 per use this month and $3 next month and $5 the month after is dying slowly, and the decline is the signal. You don’t need a single dramatic number to act; you need to notice that the denominator is shrinking while the price stays the same, because that gap is the quiet version of everything above. Flat price, falling use, month after month — the maths only ever goes one direction, and the earlier you catch it the less you’ve paid to find out.

The alternative everyone defaults to is a scheduled review — “set a reminder to check your subscriptions once a month” — and I have never found it to compete with this. A calendar reminder produces awareness without a decision: you look at the list, feel briefly guilty, and move on, because nothing on the screen is actually asking you to do anything. The per-use calculation is different because it hands you a number that either justifies the subscription or doesn’t, and a number you can’t un-see is a lot harder to ignore than a feeling you can always rationalise away. The scheduled review is how you intend to cancel; the arithmetic is how you actually do.

The honest summary is that I don’t think there’s anything sophisticated here, and that’s rather the point. Every piece of subscription advice sells you a system — a spreadsheet, an app, a calendar reminder — when the actual lever is one division you can do on the back of an envelope. Flat price divided by real use. The reason it works is that it drags a decision out of the realm of vague feeling — “this is probably too much” — and into the realm of a number you can’t argue with. And the reason it’s worth writing down rather than keeping in my head is that the flat price is always whispering that it’s fine, while the per-use price, if you’ll actually do the sum, usually isn’t. My money habits that stuck all turned out to be the embarrassingly simple ones, and this is the simplest one I keep coming back to.