Subscriptions accumulate quietly. Each one was a reasonable decision on the day it was made, and none of them announce themselves afterwards. The result, for a lot of households, is a monthly total nobody has added up and a set of services nobody has deliberately chosen in two years.
Sorting it out does not take an afternoon. It takes one honest look at what was actually watched, and a willingness to let a couple of them go.
Open each service and look at the viewing history rather than the watchlist. The watchlist is a record of intentions; the history is a record of behaviour, and they rarely match. A service with four things watched in six months is not a service you use, however good its catalogue looks when you browse it.
Do the same for every account in the house before deciding anything. A subscription that looks dead from your profile is sometimes the one somebody else uses every evening, and that is worth finding out before the cancellation rather than after.
For each service, in order: what did we watch on this in the last three months, is there something specific coming that we want, and would anyone notice tomorrow if it were gone. Most services answer themselves by the second question.
The third one is the useful test, because it separates the services that earn their place from the ones that survive on the feeling that cancelling would be a loss. Being able to resubscribe in ninety seconds is precisely why that feeling is unfounded.
It helps to add up the annual figure rather than the monthly one while you do this. Four services at a mid tier is a holiday, and seeing it as a yearly number makes the question sharper than seeing it as four small amounts leaving the account on four different dates.
Running every service at once is the expensive way to do this, and it is rarely how people actually watch. The alternative is taking one or two at a time, staying long enough to clear what you want, then moving on. A series you have waited for is usually finished in a month; three months of the service it sits on costs the same as a year of one you barely open.
This works best with a list of what you are waiting for on each service, so the decision of what to take next month is already made. Treating it as a rotation rather than a collection is the single change that brings the monthly total down without anyone watching less.
Notice how few of those are the ones subscribed to most recently. New services get used hard for a month and then settle into whatever place they really have, which is why a decision made in the first fortnight is usually the wrong one.
Write down what was still unwatched on each cancelled service and keep that note with the rest of your watchlist. It turns a cancellation into a pause: when the list has grown to three or four things, that service is worth a month again.
Check the renewal dates too. Cancelling the day after a yearly plan renews is an expensive way to be organised, and most services leave access running until the end of a period already paid for, so there is no reason to wait.
Judge each service by what it was actually used for, ask whether anyone would notice it gone, and rotate rather than stack. Two services used properly beat five that mostly sit there, and nothing is permanent either way. What quietly leaves your account each month is a regular subject over on FactMates.
