I did the arithmetic on a snack subscription box. Here is what $65 actually buys.

The box arrives on a Tuesday, which is the worst day for it, because Tuesday is when nobody in this house wants to cook and a box of foreign snacks looks like dinner.

I have now had six of them, on the middle size — thirteen to fifteen items, one box a month. That is enough boxes to have formed an opinion and enough receipts to check whether the opinion is right.

What is actually in the box

The middle size is advertised from $65 a box, and that "from" is doing real work: it is the price on a prepaid term, not the price you pay for a single month. The contents are mostly small: a bag of gummies, a packet of biscuits, a bar of something, a jar of preserve, a drink sachet, occasionally something savoury that is not a crisp.

Nine countries are represented across a typical box, which sounds like more than it is. You get one or two items from each, which is exactly the right amount for a thing you might not like.

That last point is the design of the whole product, and it is worth being explicit about: the box is built so that you cannot get too much of anything. The failure mode of a subscription is a cupboard full of the same brand; the fix is small quantities of things you have not tried.

The arithmetic, honestly

Fourteen items at $65 is $4.64 each. Let me put that next to three other numbers.

A packet of imported Japanese gummies in an Asian grocery in London is two to three pounds. A French butter biscuit tin is six to nine. A Korean honey cookie of the kind that turns up in these boxes is four or five. So the box is not a rip-off on a like-for-like basis, and it is not a bargain either.

What it is not comparable to is the supermarket. If your benchmark is a multipack of biscuits at two pounds, no subscription box will ever look reasonable, because you are not buying biscuits. You are buying fourteen decisions you did not have to make.

The number I find more interesting is the waste number. Across six boxes, roughly one item in seven went uneaten. That is not a criticism of the box — it is a criticism of how we treated it. Anything that got opened and put on a plate was finished. Anything that stayed in its packaging was still in its packaging three weeks later, and then went out.

So the real cost per item eaten was closer to $5.40, not $4.64. That is the number worth comparing, and it is the number nobody puts on the marketing page.

What the box is genuinely good at

It breaks a shopping rut. I buy the same six things from the same aisle every week and I am not going to stop. The box is the only mechanism I have found that reliably puts something unfamiliar in front of me without requiring a decision.

It is a good gift, with a caveat. Food is a safe gift because it does not require the recipient to have a shelf, a size, or a taste you can guess. The caveat is the box size — choosing thirteen items for someone else's household is a guess. The gift card route, where they pick their own plan, is the version I would actually send.

It gives you a reason to sit down. This sounds like marketing copy and it is not. The box arrived, we opened it, we put things on a plate, and we spent twenty minutes talking about a biscuit from Puglia. That is a better Tuesday than the one we would have had.

What it is bad at

Sweet things dominate. Across six boxes, the split was roughly two-thirds sweet to one-third savoury. If you are hoping for a box of interesting cheeses and pickles, this is not that box.

The packaging is a lot. Fourteen individually wrapped items, a box, a card, an outer sleeve. It is recyclable in theory and a bin's worth in practice, and any subscription of this kind has the same problem.

It creates a small obligation. A box you have paid for is a box you feel you should finish, and finishing it is not always the same as enjoying it. The cure is to give away what you do not want in the first week rather than the fourth.

Four questions that work on any food box

I have started applying the same four questions to every subscription, and they have saved me from two bad ones.

What is the price per item, not per box? A box price tells you nothing. Divide by the item count and compare that to what you would pay for one of the same thing.

How many items will you actually eat? Be honest. For us it was six in seven, which is fine. If it is four in seven, the subscription is a donation.

Is the discount a commitment? Every "as low as" price in this category is a prepaid term. Work out what you pay if you quit after two months, because that is a real scenario.

Can you give it away? A box that can be shared with a neighbour, an office, or a child's class is more useful than one that has to be consumed by the person who ordered it.

The one thing I would tell a friend

Open the box the day it arrives, put everything on a plate, and eat the strangest thing first.

That is not a joke and it is the only piece of advice I have. The box is designed so that nothing in it is precious, and the failure mode is treating it as a cupboard rather than an event. Every item that went uneaten in six boxes was one that stayed in its wrapper, and every wrapper that stayed closed stayed closed because it was easy to leave.

The second thing: buy the middle size. The small one is a week of novelty; the large one is a shelf of obligation.

Would I renew

Yes, with one change: I would move to a shorter term and keep the tins.

The reason is not that the box is exceptional. It is that the alternative — remembering to buy unfamiliar food — is a habit I do not have and am not going to build. Fourteen decisions a month for $65 is a fair trade for that, provided the plate comes out on the first night rather than the last.

If you are the kind of person who already cooks widely and shops at three different groceries, you do not need this. You are already doing the thing the box does. If you are not, and you want a low-effort way to eat something you did not choose, it is a reasonable $65 — just count the items that actually got eaten before you decide it was a bargain.