Something broke. The repair quote is on the counter and a new one is in your cart. The honest answer to "repair or replace" is arithmetic, not instinct: compare the repair price, the replacement price, and how many years each option realistically buys you. If the repair costs more than half of a good replacement and the item is already past middle age, replacing usually wins.
That half rule is a starting point, not a verdict. A $90 fix on a $300 vacuum you've owned for two years is probably worth it. The same $90 fix on a $300 vacuum you've owned for eight years is money poured into a machine already on borrowed time. The rest of this piece walks through how to run that math on shoes, vacuums, small appliances, and the awkward middle cases.
Before anything else, check one thing: the warranty. If the item is still covered, the decision makes itself — the maker fixes or replaces it at no cost. Our earlier piece on what a lifetime warranty actually covers explains why "lifetime" often means the lifetime of the part, not of you.
What is the repair actually worth per year of life?
Here's the core calculation. Divide the repair quote by the number of extra years the fix should deliver. A $120 repair that buys three more years costs $40 per year. A $400 replacement that lasts eight years costs $50 per year. On those numbers, the repair wins.
The hard part is estimating the extra years, and nobody hands you that figure. You have to reason from the failure itself. A snapped belt, a worn hose, a loose seam — these are wear items. Fixing them restores the machine to near-full health, and the rest of the device is fine. A burnt-out motor, a cracked housing, or a failed control board in an older unit is different. When the expensive heart of the machine fails, the other parts are usually the same age and not far behind.
Ask the repairer one direct question: "If you fix this, what's likely to go next?" A good technician will tell you. If the honest answer is "the motor, probably within a year or two," the repair quote is really a down payment on a replacement.
When does repairing shoes and clothing make sense?
Wearables are the friendliest repair category, because the fix is cheap relative to the item's remaining life. Resoling a well-made pair of leather shoes typically costs a fraction of a comparable new pair and can add years of wear. A tailor charging a modest fee to replace a zip, take in a seam, or reattach a strap is almost always cheaper than buying equivalent quality new.
The test is the base garment. Repair works when the fabric, leather, or construction is sound and the failure is a single component. It fails when the material itself is worn through — thinning soles on budget shoes, fabric pilling or wearing thin at friction points. Putting a new heel on a shoe whose upper is already tired just moves the breakdown date by a season.
Cost per wear is the right lens here, with the caveats we covered in our piece on the cost-per-wear formula and where it fails. A resole that extends a $200 pair from two years to five turns a high cost-per-wear into a low one. The same resole on a $40 pair rarely pencils out, because the cheap pair was never built to be resoled at all.
Is it worth fixing a vacuum or small appliance?
Small appliances sit at the opposite end. Manufacturers price many of them as semi-disposable, and parts availability is the tell. If you cannot buy the replacement part yourself — or the maker sells it only to authorized shops at a price close to a new unit — the repair market has already answered your question for you.
Run the same two-year test. Get the quote, including any diagnostic fee, shipping, and labor. Compare it against the price of a replacement with a real warranty. Then check the failure type: a clogged filter, a broken brush roll, a cracked dustbin lid are cheap, sensible fixes. A dead battery pack sealed into the body, a failed circuit board, or a burnt motor on a machine out of warranty usually tips toward replacement.
One more check before you replace: the secondhand market. Our guide to checking resale prices before you buy retail applies in reverse here too. If your broken machine still sells used for a meaningful sum even in faulty condition — some models have strong parts value — selling it for parts offsets the replacement cost. If it's worth nothing, scrap it and move on. This connects to our earlier piece, Check resale prices before you buy retail.
What about the cheap version versus fixing the good one?
Sometimes the replacement isn't the same item at all. Your mid-range machine dies and you're tempted by a budget model at half the price. That can be the right call — we've argued before that the cheap version is sometimes the smarter purchase. But compare honestly. If the budget model lasts half as long, the annual cost is the same and you've traded convenience for nothing.
The quality signals worth checking are the same whether you're repairing or replacing: metal where the stress lands, replaceable parts listed in the manual, a warranty the maker honours rather than one buried in exclusions. Our sweater comparison walks through the quality signals that separate a $40 from a $140 garment, and the same logic — materials, construction, serviceability — carries over to gear.
What this means in practice: don't replace like-for-like on autopilot. If you repair, repair the durable version. If you replace, replace upward in durability or downward in price — never sideways into the same mediocre item at today's higher price.
How do the pros run this decision?
It's worth seeing how businesses handle the same trade-off, because they do it with spreadsheets instead of hunches. Industrial buyers face it constantly: a motor controller fails, and the choice is a specialist repair or a new unit. The listed price of a new replacement makes the stakes visible. The parts seller Industrial Automation Co lists a new 75 HP Invertek Optidrive P2 variable frequency drive at $5,041.00 with a two-year warranty and four units in stock. At that scale, a repair quote of even a few hundred pounds is obviously worth it — and the two-year warranty on the new unit is part of what the buyer is paying for.
Scale that logic down. The ratio matters more than the absolute numbers. A repair under roughly a quarter of the replacement price, on an item with most of its life left, is close to a free win. A repair over half the replacement price, on an item already deep into its expected lifespan, rarely makes sense. Between those poles, the failure type and the warranty decide it.
One caution the sources above don't settle: expected lifespans for any specific product vary widely by use and model, and no figure in this piece should be read as a promise for yours. The framework holds; the inputs are yours to gather.
Practical steps: the five-question check before you pay anyone
- Warranty first. Check the purchase date and the warranty terms before anything else. A covered repair changes every number that follows.
- Get the full repair cost. Diagnostic fee, parts, labour, shipping, and whether the work carries any guarantee of its own. A repair without a workmanship guarantee is a gamble with your own money.
- Price the true replacement. Not the sticker you remember — today's price, from a current listing. Our piece on whether a "was" price is real applies to your own memory too: it drifts.
- Estimate remaining life. Ask the technician what fails next, and check whether spare parts exist at sane prices. No parts means no future repairs, which shortens whatever life a fix can buy.
- Divide and compare. Repair cost divided by added years against replacement cost divided by expected years. Pick the lower number unless the failure type says the estimate is fantasy.
Prices change after the date you check them, and repair quotes can shift once a technician opens the machine and finds more damage. Get the quote in writing and treat it as valid only while it stands.
The honest limits of this framework
The arithmetic settles most cases, but not all three. Safety is the first override: anything that overheats, sparks, leaks, or guards a blade badly gets repaired only by a qualified professional or not at all. No per-year cost justifies keeping a hazard in the kitchen. The second limit is downtime — if you need the item working this week and the repair queue is three weeks long, the math loses to the calendar. The third is sentiment, which is real but should be named as such: keeping a repaired item because you like it is a legitimate choice, as long as you know that's what you're buying.
What the evidence here establishes is the method: compare dated quotes, estimate remaining life honestly, and let the ratio decide. What it can't establish is the lifespan of your particular machine — that you learn from the technician's answer, the manual's parts list, and, eventually, the next breakdown.