The Subscription Audit Everyone Should Run Once a Year

Subscription pricing works because individual amounts are small enough to approve without thought and recurring enough to compound substantially. The result is a set of charges that most households have not consciously evaluated in years.

Start from the statement, not from memory

Listing subscriptions from recall reliably misses several. Work backwards through twelve months of bank and card statements instead, and note every recurring charge. Almost everyone finds at least one they had forgotten entirely, and frequently one for a service they believed they had cancelled.

Convert to annual figures

A monthly charge is designed to feel negligible. Multiplying by twelve changes the evaluation considerably — a modest monthly amount becomes a number that invites comparison with what else it could buy. This single arithmetic step drives most cancellations.

The relevant question is usage, not value

People defend subscriptions on the basis that the service is good. The correct test is how often it was used in the last three months. A genuinely excellent service used twice a year is still a poor purchase. Where usage data is available in the app, check it rather than estimating.

Watch for silent price increases

Subscription prices rise, frequently with minimal notice. A service signed up for at one price may now cost considerably more, and the increase is easy to miss because the charge simply continues. Compare current charges against what you remember agreeing to.

Cancellation friction is deliberate

Services that take one click to join and require a phone call to leave have designed it that way. Budget time for this, and note that many services offer a substantial retention discount when you attempt to cancel — worth knowing before you decide.

Pause rather than cancel where possible

Several services allow suspension. For anything seasonal — a streaming service you use in one part of the year, a gym during travel — pausing captures most of the saving without the friction of resubscribing.

Run it once a year, from the statements. The finding is almost always larger than expected.