A real discount with a commitment attached
Is Subscribe and Save worth it is a better question than most discount questions, because the answer is frequently yes and the failure mode is specific rather than general.
A recurring-delivery discount is a genuine reduction at checkout. That already separates it from a crossed-out reference, from cashback, and from most of what looks like a saving. Nobody is measuring it against a number you never paid.
What it comes attached to is a commitment, and the commitment is where the arithmetic gets interesting. This sits inside the pillar on judging whether a price is good.
Three numbers decide it
The per-unit price after the discount, compared with the same product bought once at what it normally sells for. Not compared with the listing's reference price, and not with the subscription's own "regular" figure.
The tier. These schemes usually pay a higher rate when several subscriptions deliver together, and that extra rate is conditional on the other subscriptions still existing. Price the purchase at the base rate and treat the tier as a bonus, because the tier is the part most likely to disappear when you cancel something unrelated.
The quantity you will actually use in the delivery period. This is the number people get wrong, and it is the one that decides whether the discount is a saving or a cupboard.
Multiply those out and compare with the alternative: buying the same item when you need it, at whatever it costs then. On a genuine staple the subscription usually wins. On anything consumed unpredictably it frequently does not, for reasons that have nothing to do with the percentage.
Where it reliably works
Three conditions, and all three have to hold.
Predictable consumption. Something you get through at a rate you could state. Detergent, coffee, pet food, filters, razor cartridges.
A stable price. Products that do not swing much, so the discount is measured against a steady base rather than against a figure that will be lower next month anyway.
Something you would buy at full price regardless. The test that catches most of the failures. If the subscription is the reason you are buying it, the discount created the purchase rather than reducing it.
Where all three hold, this is one of the better offers in retail: a real percentage off something you were always going to buy, arriving without you having to think about it. The convenience is worth something on its own and is frequently the larger benefit.
The two ways it goes wrong
Subscribing to something consumed unpredictably. The discount is real and the surplus is larger. Deliveries arrive on schedule whether or not you need them, and the accumulation is silent until the cupboard is full. This is the most common failure and it is not about the price at all.
The per-unit price that was never good. A household staple on subscription at a rate worse than a supermarket multipack, where the discount only brings it back to par. Because the comparison presented is against the same listing's undiscounted price, the external comparison never happens.
A third, smaller one: assuming the rate is fixed. The discount is a percentage off whatever the item costs at each delivery rather than a locked price, so a rate that was good at signup is not guaranteed. Checking the charge on a delivery occasionally is worth more than reading the terms.
The commitment, priced honestly
Cancelling is usually easy. Noticing that you should is the hard part, and that asymmetry is what the discount is paying for.
Two habits make it cheap. Set a reminder to review subscriptions on a schedule rather than intending to remember, and check the actual charge rather than the rate you were quoted. Most services publish enough history to make both a two-minute job.
It is also worth being clear about what kind of offer this is, because it sits between two others. Unlike a delayed benefit, it reduces the price today, which puts it on the right side of the line drawn in cashback against an instant discount. Unlike a multi-buy, it does not require you to take the quantity all at once, which is a genuine advantage on anything perishable.
What it shares with a multi-buy is the important condition: the offer is only a saving on the quantity you will actually use. The arithmetic is the same one, and the discipline is to decide the quantity before looking at the rate.
Where it is better than the alternatives
It is worth naming what this format does well, because the failure modes above can read as an argument against it.
Against a multi-buy, it wins on anything perishable or bulky. Buy one get one free requires you to take both now; a subscription spreads the same quantity over months, which removes the storage cost and the spoilage risk entirely. For a household staple that is a meaningful difference and it rarely gets mentioned.
Against buying as needed, it wins on attention. A recurring item you never have to think about, reorder or run out of is worth something beyond the discount, and for anything genuinely routine that convenience is frequently the larger half of the value.
Against a stockpile bought on a one-off deal, it wins on price risk. A subscription tracks the current price rather than locking in a moment, which cuts both ways but means you are never holding six months of something that has since become cheaper.
Where it loses is against simply not buying the thing, which is the comparison the format never presents.
Before you subscribe
Four checks, about ninety seconds.
Compare the per-unit price with the cheapest reasonable alternative you would actually buy, not with the listing's own undiscounted figure. Work out what quantity arrives per year at the chosen frequency and ask whether you will use it. Note whether the quoted rate depends on a tier, and price it without one. And check the delivery frequency options, because the difference between monthly and every two months is usually the whole decision.
Then, if the answer is yes, use the lower frequency. It is easier to bring a delivery forward when you run out than to unwind a surplus that took four months to build, and running out briefly costs less than storing three months of something you stopped using. The coupon and promotion layer that sometimes stacks on top of these rates is covered in how Amazon coupons work.
One last observation. The strongest argument for these schemes is not the discount at all, it is that they remove a recurring decision. Anything you have decided once and will keep deciding the same way is a good candidate; anything you would want to reconsider each time is not. Framed that way the answer is usually obvious within a few seconds, and the percentage stops being the thing you are evaluating.
And if you already have several running, the most valuable ten minutes available is reviewing them together rather than individually. Tiered rates, overlapping deliveries and items you stopped using are all visible in aggregate and invisible one subscription at a time, which is precisely the asymmetry the format relies on.
The shortest useful test, if you want one: would you set a reminder to buy this item on this schedule if there were no discount at all? Where the answer is yes, the subscription is doing something you wanted anyway and the discount is a bonus. Where it is no, the schedule is the product and the discount is what is paying for your agreement to it.
