There is a whole class of products where the machine is close to a giveaway and the supplies are where the money is. Printers are the famous case, but the pattern is everywhere: the coffee machine that only takes its own pods, the vacuum with proprietary bags, the water filter jug, the razor. None of this is a scandal. It is a pricing structure, and it works because the number on the shelf edge is the only one most people compare.
The alternative takes about ten minutes and needs no expertise. You are adding six numbers instead of one, then dividing by how long the thing will realistically serve you. What comes out is a cost per year or per use, which is the only figure that lets you compare two genuinely different options.
| Purchase | What you hand over, plus delivery, installation and anything it needs on day one |
|---|---|
| Consumables | What it eats to do its job: cartridges, filters, pods, bags, blades, detergent |
| Energy or fuel | Rated consumption, times how often you run it, times the price on your own bill |
| Maintenance | Scheduled servicing and the parts that are designed to be replaced |
| Repair risk | What a common failure costs, weighted by how likely it is before the end of its life |
| End of life | Disposal or recycling, minus whatever it is still worth to somebody else |
Where each number actually comes from
None of these require guessing. Consumables are printed on the packaging along with a yield: pages per cartridge, liters per filter, loads per bottle. Divide price by yield and you have cost per use, which is the number that matters and never the one on the shelf. Energy comes off the rating plate or the efficiency label, multiplied by your own usage and the unit price from your own bill, which you can read straight off the statement. Service intervals are in the manual. Repair costs take one phone call to a local repairer, who will usually tell you what fails on that type of machine before you have finished asking.
The step people skip is normalization. Two options can only be compared once they are expressed in the same units: per page, per load, per mile, per year. Cartridge prices are not comparable. Cost per page is. The moment you convert, differences that were invisible on the shelf become obvious, and occasionally enormous.
The denominator nobody gets right
How long you will keep it is the number with the most influence over the result, and it is the one people are least honest about. There are three separate ways something ends: it wears out, it becomes unsupported, or you simply lose interest in it. The third one is real and it belongs in the arithmetic.
If your history says you replace a phone every couple of years regardless of condition, then a phone's theoretical decade of life is irrelevant to your calculation. If the last three of a given appliance you owned all died at roughly the same age, use that age, not the marketing. Your own replacement history is better evidence than any specification, because it includes you.
The cheapest version of something you use every day is rarely the cheapest thing you can buy.
Where the whole-life view flips the answer
Once you start doing this you notice the same handful of patterns turning up in different aisles.
- The inexpensive machine with costly supplies, where a year of ordinary use costs more than the machine did.
- The expensive machine with almost nothing to feed it, which looks extravagant on day one and stops costing anything afterward.
- The appliance that runs continuously, where energy quietly dwarfs the purchase price over its life, as against the tool that runs ten minutes a week and where the purchase price is essentially the whole story.
- The sealed, glued or unserviceable item, where the first failure is the end of the object rather than an inconvenience.
- The purchase that is cheap to acquire and expensive to keep legal, insured and roadworthy, which is most of the vehicle category.
There is a single heuristic under all of that. The more hours a year a thing runs, the more running costs dominate, and the more a higher purchase price has room to pay itself back. The less it runs, the closer true cost gets to sticker price, and the stronger the case for buying the modest version, borrowing one, or renting for the weekend you actually need it.
Running the comparison
- Fix a horizon: the number of years you will honestly keep this, not the number the manufacturer implies.
- For each candidate, total the six components across that horizon, using your own usage rate for anything that scales with use.
- Subtract a realistic resale or trade-in value, which for most household goods is small and for a few categories is not.
- Divide by years, or by uses if the thing is used in bursts. Now you have comparable numbers.
- Check whether the gap between candidates is bigger than the uncertainty in your own estimates. If it is not, the two options genuinely cost the same and you should choose on something else.
That last step keeps the exercise honest. These are estimates built on other estimates, and they are accurate to within a wide band. The purpose is not precision. It is to catch the cases where two things that look comparable on a shelf turn out to differ by a factor over their lives, which happens often enough to make the ten minutes worthwhile.
What the arithmetic leaves out
Some real costs never make it into a column. Your time, the nuisance of a machine that breaks in the middle of something, the days without a working one while a part is on order, and the plain fact that some things are a pleasure to use and some are a small daily irritation. A tool that saves twenty minutes a week is buying you something the spreadsheet cannot price.
The most useful place to try this is not a purchase at all. Run it on something you already own and were about to replace, with the repair quote as one of the candidates. The arithmetic says repair more often than instinct does, and it is the only version of this exercise where you find out afterward whether it was right.