Everybody's Subscriptions, Itemized

The guess is the interesting part.

Before any statement was opened, each household wrote down what it thought it spent a month on subscriptions. The median guess was $84. The median actual figure was $155.

Nine households. Nine underestimates. Not one household guessed high, or even close.

What they remembered

Streaming. Everybody remembers streaming, and the memory is accurate — eight of nine were within a few dollars on video and music, because those are the services a household actually uses and periodically argues about.

If your subscription problem is streaming, you do not have a subscription problem. You have four things you use and can name.

What they missed

Annual renewals. The largest single miss, in eight of nine households. A $79 charge in March is invisible in a monthly view of spending and invisible in a household’s sense of itself, and it is precisely the kind of thing that renews for six years past its usefulness. Median across the households: $34 a month in annual-equivalent charges, none of it felt as monthly spending.

Small app charges. Twenty-six dollars a month, in pieces of $2.99 and $6.99, spread across four devices. Seven households missed most of this.

Anything on a child’s account. Six households. One 12-year-old was paying $4.99 a month for something related to a game he had stopped playing in 2025; he did not know either, because the charge lands on a parent’s card and neither party ever sees it.

Duplicates. Two households were paying for the same service twice, in both cases because it had been signed up for on two different devices with two different email addresses. One of them had been doing so for nineteen months.

The unidentifiable charge

Six of nine households had at least one line nobody could explain. Not a large one — the median was $8 — and in four cases they never established what it was.

I want to sit on that for a second, because it is a stranger fact than it first appears. These are competent people with organised finances. What it demonstrates is that the household’s monitoring system for recurring payments is: somebody notices something. And nobody notices $8 — the same blindness that hides the grocery number.

Why the cull fails

Five households did what you would expect after an exercise like this and cancelled aggressively. Four months later, they had resubscribed to about half.

That is not weakness. Several of the cancellations were genuinely wrong — a service somebody depended on, cut in the enthusiasm of an evening — and the resubscription was the household correcting an error.

The four households that did something durable did something duller. They moved every subscription onto one payment method: one card, used for nothing else, so that a single statement is the complete inventory. Their spending barely changed in the first month.

What changed is the cost of looking. An audit that took an evening now takes four minutes, and a household that can audit its subscriptions in four minutes does it two or three times a year instead of once a decade. By the second audit, all four totals were down — not because anyone had a purge, but because the $8 charge nobody could identify was now sitting on a statement with eleven lines on it instead of ninety.

Make it visible, then decide. Deciding first, in a burst of irritation, on a Tuesday night, is how households end up back where they started by spring.