How to split the cost of shared subscriptions

The short version

Add up what leaves one card every month for streaming, music, storage and delivery, and the total surprises almost everyone who actually does it. A realistic bundle of four ordinary plans comes to $55.96 a month, or $671.52 a year. Nobody signed up to carry that alone. It happened because each charge is ten to seventeen dollars, automatic, and easy never to notice.

Split the bundle evenly and it barely matters whose name is on which plan: $18.66, $18.65 and $18.65 a month. Split each plan by who actually uses it, and the same money lands very differently: $28.65 on the heaviest user, $10.68 on the lightest. Both answers are defensible. The mistake is landing on either one without ever asking the question.

Why subscriptions go untracked

A subscription is the one shared cost that never has to ask anybody to notice it. Rent gets discussed once a year. A grocery run gets typed into a group chat the day it happens. A subscription renews itself, on a card nobody is looking at, for an amount small enough that checking it feels like more effort than it is worth.

That is the whole mechanism, and it has nothing to do with anybody behaving badly. The person whose card is on file did not choose to fund the household’s music. They added a card once, during a free trial, years ago, and every plan since has quietly attached itself to the same saved number. Nobody is dividing this on purpose. It is dividing itself, by whoever happened to be holding a card at the time.

• A single line looks too small to raise. $9.99 or $16.99 does not feel worth a conversation. Twelve of them a year do.
• The charge has no date and no receipt. There is nothing that arrives to remind anybody it exists, the way a bill or a group dinner does.
• Only the monthly number ever gets seen. Nobody adds twelve of them together, which is the only version of the number that changes minds.

None of this is unique to any one household. It is the default shape of every subscription, which is exactly why it is worth naming once rather than discovering it a year and a half later, when somebody finally opens a statement and asks what four plans nobody remembers agreeing to are doing on there.

One household’s bundle, split two ways

Three people, Ana, Ben and Cleo, share an apartment. Between them they pay for a film service at $15.99 a month, a family music plan at $16.99, a storage plan at $9.99 and a delivery membership at $12.99. All four sit on Ana’s card, because Ana happened to start the free trials. That is $55.96 a month, and how it should be split depends on which of two questions the household actually asks.

The film service and the storage plan barely move between the two rows, because all three of them watch about the same amount and back up about the same number of photos. The music plan and the delivery membership are where the two questions genuinely disagree: Ana logs into the music plan most days and places the delivery orders most weeks, and Cleo does neither much. Splitting everything evenly hides that difference completely. Splitting by usage puts $9.99 more on Ana every month and takes $7.97 off Cleo, for the exact same four plans.

Which of the two rows is right is really the wider question of which method fits which cost, and it has its own page: how to split shared costs fairly walks through all five ways, evenly, by income, by space, by usage, by item, and says how much each choice is actually worth arguing about before you spend an evening on it.

When one person uses the family plan for everyone

A family plan is the clearest version of this problem, because its whole pitch is that the price does not change whether one person uses it or four do. On the music plan above, Ana logs in on about 24 days a month, Ben on 10, Cleo on 2. Split the $16.99 evenly and everyone pays roughly the same, about $5.66. Split it by those days and Ana pays $11.33, Ben $4.72, Cleo $0.94.

Neither number is wrong on its own. An even split is the right answer when a family plan genuinely gets used by the family, and nobody is arguing Cleo owes for a service she signed up for and mostly ignores. What is wrong is not asking the question at all, and defaulting to evenly because dividing by three is the version that requires no conversation. $11.33 a month is $135.96 a year Ana is not currently being asked for.

The same weighting already has its own page for the bill that made the case first: splitting utility bills when usage isn’t equal works through the same logic for electricity and gas, plus the honest caveat that a usage weight is always agreed on, not measured off a meter that happens to exist for a subscription.

The tier problem: an upgrade for one becomes a bill for three

Ben wants the film service’s bigger tier for the television he just bought, so he upgrades the household’s plan from $15.99 to $22.99. Nothing else about the plan changed for Ana or Cleo; they still watch what they always watched, on the screens they always used. The $7.00 increase bought Ben something. It did not buy the household anything.

Keep the original split and add the difference to Ben alone, and Ana and Cleo still pay $5.33 each, Ben pays $12.33. Split the new $22.99 evenly instead and Ana and Cleo’s bill rises to $7.67, money they are paying toward a screen that is not theirs, while Ben’s actual cost drops to $7.66, less than half of what his own upgrade cost. Split evenly, that is $4.67 a month, $56.04 a year, moving from the two people who did not ask for the upgrade to the one who did.

The reason this one is easy to miss is that the upgrade never shows up as a new line. The plan name does not change, the payment does not change, and the only evidence is a bigger number on a charge that already looked routine. Ana and Cleo have no reason to check whether $22.99 is still a $15.99 plan split evenly with an extra $2.34 hidden inside it, because nothing about the bill announces that the question changed.

Check your own bundle

Put in what your household actually pays for, and mark each line evenly split or split by who uses it. The defaults below are the same four plans and the same usage numbers as the table above, so the same $28.65 and $10.68 come out the other end before you change anything.

The leaver problem: what a family plan costs when somebody moves out

A family plan’s price is set for a household, not for however many people currently live in it. When Cleo moves out, the $16.99 music plan does not get cheaper. Whatever she had been covering gets picked up by whoever is left, whether anybody decides that on purpose or not.

If the plan had been split evenly, Ana and Ben each go from $5.67 to $8.50 a month, an increase of $2.83 each, or $33.96 a year, for a seat that is not doing anything different than it did the month before. If the plan had already been split by usage, the same departure barely registers: Ana goes from $11.33 to $11.99, up 66 cents, because Cleo’s $0.94 share was never large enough to be missed.

Enter it once, as recurring, not twelve times

A subscription is the cleanest case there is for a recurring expense rather than a repeated one. The amount does not change, the day does not move, and the split, once agreed, does not need re-deciding every month. Logging $16.99 by hand on the same day for a year is twelve chances to forget once, and forgetting once is enough to make the running total wrong for good.

Set it up the way the charge actually behaves: an amount, a monthly cadence, and whichever split the household settled on for that particular plan, evenly for the film service, by usage for the music plan. From there it posts itself, and the number that mattered all along, the twelve-month total nobody adds up in their head, is just sitting there instead of waiting to be discovered.

A recurring rule also gives a price rise somewhere to land. Subscriptions go up a dollar or two most years, quietly, and a repeated manual entry just keeps logging the old figure until somebody notices the card statement does not match. A rule that renews itself shows the new amount the next time it posts, which is one more reason the shortfall in the tier problem above goes unnoticed for so long: a price change and an upgrade look identical from the outside, a bigger number on the same line, and neither one announces which of the two just happened.

The extra cent that landed on Ana in the evenly split row above, $18.66 against $18.65 twice, is not an accident of this particular example. Where the extra cent goes explains why the rule that places it is deterministic rather than arbitrary, and why the same household gets the same answer, on the same plan, every single month.

Common questions

Every figure above came from running Halvy’s own split engine on the worked household rather than from arithmetic done by hand: the four plans reconcile to $55.96 exactly whichever way they are split, and the tier and leaver figures were checked the same way, including a direct check that the base split plus the upgrade increment sums to the new plan price to the cent.