2 September 2026 · Guide · 7 min read

How to audit which streaming services you actually watch

Most advice about cutting streaming costs stops at "cancel what you don't use", which is useless, because nobody knows what they don't use. This is a method that produces a number you can act on, in about forty-five minutes.

Step 1 — Find every charge, not the ones you remember

Start from the money, never from memory. Open twelve months of statements for every card and account that could plausibly carry a subscription and search the text for the service names you can think of, then read the rest of the lines anyway. The ones people miss are never the big obvious service.

Four places charges hide:

App store billing. Subscriptions bought inside an app appear on your statement under the platform's name, not the service's, so a search for the service name will not find them. Open the subscriptions list in your phone's account settings and read it line by line.
Add-on channels. Several large services resell other people's channels inside their own app. The charge shows up as one line from the host service even though you are paying for two things.
Bundles. Phone carriers, broadband providers and pay-TV packages fold streaming into a single bill. These are the hardest to price honestly, and step 2 deals with them.
Annual plans. A yearly charge appears once and is easy to skim past in a monthly review. It is also the one you are least likely to be actively using, because you decided about it a long time ago.

Write down every line with the amount and the billing date. Do not judge anything yet.

Step 2 — Convert everything to one annual number

Comparison only works when the units match. Multiply monthly prices by twelve. Take annual prices as they are. Include tax exactly as it appears on the bill rather than the advertised price, because the advertised price is not what leaves your account.

Two rules that keep this honest:

Count add-ons as their own line. If a channel is billed through another service, it still gets its own row. Otherwise the host service inherits viewing it did not earn and looks better than it is.
Price bundles by the alternative. A bundle has no honest per-service price. What it has is a counterfactual: what would the bill be if you dropped the streaming component? If the provider will not let you drop it, the streaming is genuinely free at the margin and belongs at zero — but only until the bundle renews, at which point it becomes a real number again.

Add the column up. That total is the figure the rest of the audit is about, and it is usually larger than the guess you would have made.

Step 3 — Count what you actually finished

Now the harder half. You need viewing data, and the good news is that every service keeps it for you.

Most services expose a viewing activity or watch history page in account settings — usually easier to reach in a web browser than on the television. Where there is no history page, the continue-watching row and the "my list" section on the TV are a reasonable substitute, and so is the profile's recommendations page, which is built from what you watched.

For each service, list the titles you finished in the last twelve months. A film watched to the end counts as one. A series watched to the end of a season counts as one. Three episodes and then a stop counts as zero — you did not get the thing you were paying for, and counting it flatters the service.

Count titles, not hours. Hours reward whichever service you leave on in the background, which is the opposite of what you are trying to measure. You are measuring how often a service was the reason you sat down.

Do it per household, not per person, and use every profile. A service kept for one person watching one show is a completely legitimate answer — but you want to know that is what it is.

Step 4 — Divide, and read the result carefully

For each service: annual cost divided by titles finished. That is your cost per finished title.

The number will be uncomfortable at the bottom of the list, and that is the point. A service you finished one thing on over a year costs you its entire annual price for that one thing. Whether that is bad depends on how much you wanted it, but you should be making that trade knowingly rather than by direct debit.

Three honest caveats before you act on the ranking:

It is a ranking device, not an accounting fact. Its job is to sort the list, not to produce a defensible per-title price.
Seasonality distorts it. A service you use heavily for two months of the year and never otherwise will score badly on an annual view. That is an argument for subscribing in bursts, not necessarily for cancelling.
Live sport and news break the model entirely. There is no finished title. Substitute events attended or a rough count of sessions and keep it in its own group.

Step 5 — Price the alternative before you cancel

Take the worst line. Ask what would actually have happened if it had not been there: which two or three titles would you have missed, and what would they cost to rent or buy individually? Look the current prices up at the moment you do this, because rental and purchase prices move, and any figure written into a guide like this one would be wrong by the time you read it.

If a year of the subscription costs more than renting everything you finished on it, the subscription is a bad deal on its own terms and the maths is doing its job. If it costs less, you have just proved the subscription is worth keeping, which is also a useful result.

The second alternative is timing. Nearly every service can be cancelled and resubscribed later without penalty, so a service you use in bursts is often better bought in bursts. The cost of that is friction and remembering — which is the problem a watchlist that survives across services is there to solve.

Step 6 — Cancel, diarise, then check again in ninety days

Cancel at the end of the paid period rather than immediately; you have already paid for the remaining days. Note two dates: when access ends, and a reminder ninety days out.

At ninety days, re-run the count on what is left. Two things show up. First, whether you quietly resubscribed to something out of habit. Second, whether the viewing you lost simply moved to another service — which means the cancellation worked and the money is now doing more.

Where JumpIn fits, and where it does not. JumpIn is an Apple TV app that searches your services once and shows which of them carry a title, with what it costs, before you press anything. It never plays video, and it cannot see your bank statements — the money part of this audit is yours to do.

What it can do is show, on your own television, how often a given service is actually the answer to "where is this", and tell you when the thing you want is cheaper to rent than to subscribe to. That data stays on your device. Availability and prices are reported from catalogue data that changes constantly, so treat what any tool shows you — ours included — as current rather than permanent. JumpIn is not on the App Store yet; join the waitlist and we will tell you when it is.