The Subscription Audit

A subscription is priced monthly because $12.99 does not feel like $155.88.

That is not a criticism of the pricing. It is a description of why it works — and the entire point of an audit is to convert every recurring charge back into the number you would have reacted to if you had been shown it up front.

Nine subscriptions at $12.99 is $1,403 a year. That is not a hypothetical figure chosen to shock; it is close to what an ordinary household's recurring digital and service charges add up to, and it is almost exactly the size of a typical annual-bills sinking fund.

A note before you start. This is general education, not financial advice. The subscription prices used here are illustrative, and the sinking-fund figures are computed by this site's own savings-goal engine.

1. The arithmetic the monthly price hides

The five-year column is the honest one for anything you have not actively reconsidered, because a subscription nobody cancels is a subscription you are buying for five years.

A single $19.99 service kept for five years is $1,199. That is a real number and it competes with real things.

And the totals compound faster than intuition suggests:

2. What the money is competing with

This is the framing that makes an audit actionable, rather than an exercise in guilt.

Cancelling $65 a month is $780 a year. Here is what $780 a year is, in the language of the rest of this site:

4.3% of an $18,000 five-year car sinking fund, which needs $300 a month. The cancellation funds two and a half months of it.

Most of a $1,400 annual-bills sinking fund, which needs $117 a month — the audit covers 56% of it outright.

Or, if it goes into an emergency fund at $65 a month, roughly two and a half months of runway added over three years for a household whose essential spending is $2,400 a month.

The point is not that subscriptions are bad. It is that $65 a month is a real allocation decision that was never actually made — it accumulated one $12.99 at a time, and each individual decision was defensible.

An audit is simply the moment where the decision gets made once, deliberately, with the annual figure visible.

3. The test that works

"Do I use it?" is the wrong question, because the answer is almost always technically yes, and technically-yes keeps everything.

The question that works: would I buy this again today, at today's price, if I did not already have it?

Three reasons that version works better:

It removes the sunk cost. What you have already paid is irrelevant to whether the next year is worth it, and the "do I use it" framing quietly smuggles it back in.

It reprices at the current price. Many subscriptions cost more now than when you signed up — see section 4 — and the question forces you to evaluate the price you are actually paying rather than the one you agreed to.

And it produces a decision rather than a feeling. "Yes" and "no" are both complete answers. "I use it sometimes" is not.

A useful secondary test for the borderline cases: cancel it and see whether you miss it. Almost everything can be resubscribed, usually within a minute, and the cost of being wrong is one month of a service you were paying for anyway. The asymmetry favours cancelling.

4. The three mechanics that do the damage

Most subscription creep is not carelessness. It is three specific mechanisms, and each has a specific defence.

Free-trial conversion

A trial that requires a payment method converts by default. The business model assumes a meaningful share of people will not cancel in time, and it is right.

The defence is a calendar reminder set at the moment you start the trial, dated two days before it converts — not the day it converts. Do this as part of signing up, not afterwards.

Silent price increases

A subscription you signed up to at $9.99 may now be $15.99. The increase was disclosed, usually by email, usually once, and almost nobody re-evaluates on receiving it.

The defence is the audit itself. You cannot detect this without looking at the current charge, which is why the audit works from your bank or card statement rather than from memory. Memory holds the price you signed up at.

Annual renewals you forgot were annual

These are the most expensive ones and the hardest to see, because they generate a single large charge once a year and then eleven months of silence.

Domain names, cloud storage, professional memberships, software licences, warranties, roadside assistance, subscription boxes on a yearly cycle.

The defence is that the audit must cover twelve months of statements, not one. A one-month review is structurally blind to the entire category.

5. How to actually run it

Forty-five minutes, once a year. The method matters more than the effort.

Pull twelve months of statements — every current account and every card, including any you rarely use. Twelve months, because of section 4.

Mark every recurring charge. Do not judge yet; just list. Include the ones you are certain about, because certainty is exactly where the forgotten price increases hide.

Write the annual figure next to each one. Monthly × 12. This step is the whole audit — everything before it is data collection and everything after it is decision-making.

Total the column. That number is your real subscription spend, and for most households it is meaningfully larger than their estimate.

Then apply the section 3 test to each line, in descending order of annual cost. The order matters: starting with the biggest annual figures means the audit is worth doing even if you lose interest a third of the way down.

And decide where the cancelled money goes, in the same sitting. Money freed and not reallocated does not stay freed — it is absorbed within about two months, and next year's audit finds the same total.

6. Where they actually hide

The audit finds what you look for, so the categories matter. Six places recurring charges accumulate, roughly in order of how often they are missed.

Streaming and media. The category everyone thinks of, and rarely the largest. Video, music, audiobooks, news, and the add-on channels bought inside another service, which appear on the statement under the parent service's name and are invisible.

Software and cloud. Storage that outgrew its free tier, a design tool bought for one project, password managers, VPNs, backup. Frequently annual, so frequently missed.

Health and fitness. Gyms, classes, apps, meal plans. Gym contracts are the single most common example of a subscription that continues after usage stops, and often the hardest to cancel by design.

Delivery and retail memberships. Free-shipping programmes, subscription boxes, loyalty tiers, grocery delivery. These are the ones most likely to have had a silent price increase, because they are positioned as saving money.

Insurance and protection add-ons. Device protection, extended warranties, roadside assistance, identity monitoring. Often bought once at a point of sale and never revisited, and often duplicating cover you already have — a credit card or home insurance policy may already include several of them.

Professional and financial. Memberships, journals, certifications, advisory fees, account maintenance charges. The category most likely to be genuinely worth keeping, and worth checking anyway, because dues rise and roles change.

Two mechanical hiding places worth naming separately:

Charges billed through an app store appear under the platform's name rather than the service's, so a statement line reading only the platform tells you nothing. You have to open the subscriptions list inside the account itself.

And charges on a card you rarely use — an old card, a secondary card, a card kept for one purpose — will not appear on the statement you actually review. Section 5 says every account and every card for exactly this reason.

7. Two things worth keeping

An audit that cancels everything has failed as badly as one that cancels nothing.

Keep what you would re-buy. If the answer to section 3's question is a clear yes, the subscription is doing its job and there is nothing to fix. A budget with no discretionary spending in it does not survive.

And keep the ones that replace something more expensive. A service that genuinely substitutes for a larger cost — transport, tools you would otherwise buy outright, software your income depends on — is not the same category as entertainment, even though it looks identical on a statement.

One thing worth adding rather than removing: switch the survivors to annual billing where it is offered at a discount — but only after they survive the audit. Annual billing on something you would have cancelled is a twelve-month commitment to a mistake, which is why the order is audit first, switch second.

And put next year's audit in the calendar before you close the spreadsheet. This is a recurring exercise, not a one-time cleanup, because the three mechanics in section 4 start working again the day you finish.

8. Where this sits in a budget

Subscriptions are the easiest line to act on and among the smallest. Both halves of that sentence matter.

Easiest, because it is the only major budget line you can change this afternoon. Housing takes a lease cycle. Transport takes a sale. A subscription takes four clicks, which is why an audit is the standard first move when a budget does not balance.

And among the smallest, because $780 a year is real and it is not a housing decision. A household whose rent takes 80% of its needs bucket — which describes eighteen states at $45,000 — will not fix that with a subscription audit, and telling them otherwise is the kind of advice that makes budgeting feel like a moral test rather than arithmetic.

Do the audit because it is fast, free and repeatable. Do not mistake it for the answer to a structural shortfall.

9. Cancelling is designed to be hard, and that is a signal

A service that makes cancellation difficult has told you something about how it expects to retain you, and it is worth reading that signal rather than being worn down by it.

The common friction patterns:

Cancellation only by phone, during business hours, when signup took thirty seconds online. The asymmetry is the design.

A retention flow with three or four screens, each offering a discount, a pause, or a downgrade. A discount offered at cancellation is a disclosure that you were paying above the price they will accept — which is worth knowing whether or not you take it.

A "pause" offered instead of cancellation. Sometimes genuinely useful. Often it is a cancellation that resumes automatically, so check the resume date and set a reminder for it.

And cancellation that takes effect at the end of a term you have already paid for. This is fair and it is worth confirming, because the service continues and it is easy to assume the cancellation did not go through.

Three practical rules:

Cancel from the account the subscription was created in, not from the card. Blocking the card is not cancelling — it can leave you owing the balance and, with a gym or a contract service, in collections.

Take the confirmation. Screenshot or email, with a date. The single most common dispute is a service that has no record of a cancellation you are certain you made.

And do it at the start of a billing period, not the end. You keep the service you already paid for and you remove the deadline pressure, which is what the retention flow is counting on.

The broader point: friction at cancellation is a cost of the subscription that you do not pay until you leave. Factoring it in at signup is reasonable — a service that is easy to leave is worth slightly more than an identical one that is not.

10. What the audit is actually measuring

One reframing worth ending on, because it changes what you do with the result.

The total you produce in section 5 is not a measure of waste. It is a measure of decisions made by default.

Some of those defaults will be correct. A service you would re-buy at today's price is a good decision that happens to have been made passively. Nothing needs to change.

Some will be stale rather than wrong. A subscription that was right when you signed up and is not now — a gym near an old flat, software for a job you left, a service whose price doubled. The decision was sound and the circumstances moved.

And some were never decided at all. A trial that converted. A tier that upgraded. A charge you did not recognise when you saw it on the statement.

The three categories want different responses, and lumping them together as "waste" produces the guilt reaction that makes people avoid doing this at all.

The useful annual question is not "how much am I wasting" but "how much of my spending is on autopilot" — and the answer for most households is a few thousand dollars a year across subscriptions, insurance add-ons and memberships combined.

That number is not shameful. It is normal, it is what recurring billing is designed to produce, and it is entirely fixable in forty-five minutes once a year.

Frequently asked questions

How much do subscriptions actually cost? More than the monthly figure suggests, which is the point of monthly pricing. A $12.99 service is $155.88 a year and $779 over five. Nine of them is $1,403 a year.

What is the best test for whether to keep one? Not "do I use it" — that answer is almost always technically yes. Ask whether you would buy it again today, at today's price, if you did not already have it. That removes the sunk cost and reprices at the current rate.

Why do I need twelve months of statements? Because annual renewals generate one charge and then eleven months of silence. A one-month review is structurally blind to domain names, professional memberships, warranties and anything else on a yearly cycle — which are often the most expensive items.

How do I stop free trials converting? Set a calendar reminder for two days before the trial ends, at the moment you sign up rather than afterwards. The conversion rate on unattended trials is exactly what the pricing model assumes.

Should I switch to annual billing to save money? Only after a subscription has survived the audit. Annual billing on something you would otherwise have cancelled is a twelve-month commitment to a mistake.

What should I do with the money I free up? Decide in the same sitting, and move it automatically. Money freed and not reallocated is absorbed within about two months, and next year's audit finds the same total. $65 a month covers 56% of a typical $1,400 annual-bills sinking fund.

Is a subscription audit going to fix my budget? If the problem is leakage, yes — it is the fastest line to change and you can do it today. If rent is taking 80% of your needs bucket, no. Cancelling $780 a year does not close a structural gap, and treating it as though it should is how budgeting becomes a moral test.

How often should I do this? Annually, on a calendar reminder set before you finish the current one. Price increases and trial conversions start working again immediately.

What to do next

Pull twelve months of statements and write the annual figure next to every recurring charge. That single column is the audit.

• 50/30/20 budget calculator — where the freed money should go.
• Sinking funds — the destination an audit funds outright.
• Zero-based budgeting — the method that stops the leak returning.
• The 50/30/20 rule and where it breaks — when the problem is not subscriptions.

Every figure on this site is sourced and dated. How we source every number.

Subscription prices used throughout are illustrative round numbers chosen to demonstrate the arithmetic, not measured market averages — the annual and five-year conversions of them are exact. Sinking-fund contribution figures are computed by this site's own savings-goal engine. Emergency fund runway figures assume the stated essential monthly spending and no investment return. This is general education and not financial advice.