For most purchases, no — and the reason is structural rather than cynical. An extended warranty is insurance, and the price of any insurance has to cover the repairs the seller expects to pay for, plus administration, plus the commission of whoever sold it to you, plus profit. That arithmetic means the average buyer gets back less than they paid. Insurance is still worth buying when the loss would be one you genuinely cannot absorb. So the question is not "will this thing break?" but "if it broke tomorrow and nobody helped, would I be fine?"
That test settles most of these offers in about ten seconds, and it explains the familiar advice: insure the house, decline the plan on the headphones. What follows is the structure of the cost — what pushes the price of a protection plan up and down, and why the offer arrives at the moment you are least equipped to evaluate it.
What an extended warranty actually is
Three different things get called a warranty, and telling them apart decides the answer.
- The manufacturer's warranty included with the product: already paid for inside the price, and typically covering defects — the product failing to be what it was sold as — not accidents or wear.
- Statutory or consumer-law protection, which exists in many countries and varies enormously in strength and length. These are rights rather than a product, and in some places they overlap heavily with what a paid plan sells back to you.
- The extended warranty or protection plan: an optional contract you pay for separately, extending the term and sometimes widening the coverage to accidental damage or no-fault replacement.
Only the third is a decision. And because it pays out on an uncertain future event in exchange for a certain payment now, it is insurance, whatever the packaging calls it.
Why the price has to exceed what the average buyer gets back
The price of insurance is built from the same parts every time: expected claims across everyone in the pool, the cost of running the claims process, the sales commission, and a margin. In the language of expected value — the average result of a bet, weighing each outcome by how likely it is — that makes buying insurance negative expected value for the buyer by design. Not a scandal waiting to be exposed; it is the fee for moving a risk off your balance sheet and onto someone else's, and it is worth paying when the risk is the wrong shape for you to carry.
What makes point-of-sale plans harder deals than most insurance is the commission line. Coverage sold face to face, in the last thirty seconds of a purchase, carries the cost of being sold that way. Coverage already attached to a payment method or a household policy does not.
The test that decides it: could you absorb the loss?
Expected value tells you the average buyer loses. It does not tell you to decline, because you are not deciding about averages — you are deciding about your own survivable range. That boundary is the entire argument for owning insurance at all, and the useful question is the size of the loss relative to your buffer.
- Losses you can absorb without changing anything — a kettle, headphones, a cheap monitor. Self-insure. Paying a margin so someone else handles a cost you could meet from petty cash is an expensive convenience.
- Losses that would sting but not break you — a laptop, a phone, a washing machine, for most households with savings. Usually decline; this is the honest grey area where the specifics below matter.
- Losses you could not absorb — borrowing at a bad rate, missing rent, losing your ability to earn. Insure, even at a poor expected value, because that outcome takes you out of the game rather than merely setting you back.
Notice what the test ignores: the probability of failure. Probability matters for pricing, but you are not the one pricing it. Size of loss against buffer is the part you actually know. The general version of this move — matching the effort of a decision to what it can cost you — is in our framework for making better decisions.
What drives the cost of a protection plan up and down
Any quoted figure is specific to the item, the seller and the term, so the useful thing to know is what moves it. Both the price of a plan and the value in it turn on:
Repair economics. Sealed assemblies, glued batteries, bonded screens and refrigerant systems make repairs expensive and sometimes uneconomic. The more likely a failure means replacement rather than repair, the more the plan costs — and the more it can be worth, if it swaps rather than repairs.
Term length, and how much of it is duplicate. The headline is total years, but months overlapping the included manufacturer's warranty are months you pay for twice. The honest price is per year of additional cover, and the honest comparison between two plans uses that.
Excess and claim limits. An excess per claim, a cap per claim, or a lifetime cap on the contract all shrink the payout while the premium stays where it is. A large excess on a mid-priced item can leave you insured against a loss you could already absorb.
Exclusions. Accidental damage, liquid, wear, batteries, cosmetic damage and commercial use are the usual carve-outs, and between them they cover most of the ways things actually stop working. Coverage that excludes your realistic failure mode is not cheap; it is irrelevant.
Who backs it, and how a claim runs. Manufacturer plans, retailer plans and third-party administrators differ in whether you end up with a repair, a refurbished unit, a store credit, or an argument — and turnaround time is part of the product. The same coverage is also priced differently bundled into finance, offered at the till, or bought standalone, and the till is the most expensive counter in the shop.
Why the offer feels most persuasive exactly when it should not
The pitch arrives after you have chosen, while you are holding the thing, in a conversation you would like to end. Three documented forces point the same way at once.
Anchoring — the first number in view setting the scale for the numbers after it — makes the plan feel small. Against an expensive item it is "only a fraction of the price"; against your monthly budget, which is the comparison that matters, it looks like something else entirely.
Loss aversion, documented by Daniel Kahneman and Amos Tversky, is the finding that losses loom larger than equivalent gains. The salesperson needs no argument: describing a cracked screen is enough, because an imagined loss carries more weight than the certain, smaller loss of the premium. Add the fatigue of three decisions already made in the last twenty minutes, and the fourth gets very little scrutiny — which is why it is placed there. The whole family of these effects is in our field guide to cognitive biases.
Vividness beating frequency. One story about a friend's failed device outweighs the duller question of how often that model needs repair across everyone who owns one — the reflex in our piece on thinking in base rates. Your own history is the better base rate: two broken phones in five years is real evidence about you, and it argues specifically for accidental-damage cover.
The counter is not willpower but a policy set before you are in the shop: I insure what I could not replace, and decline the rest. A rule set in advance is honest; a rule improvised while someone waits for an answer bends toward whoever is talking.
When an extended warranty does earn its price
The honest cases share a shape — either the loss is unaffordable, or the coverage does something your baseline does not.
- No buffer right now. If replacing the item would mean borrowing expensively, the plan buys certainty you genuinely lack. A real reason, and one that fades as savings grow.
- The tool you earn with. A freelancer's laptop, where a week without it costs more than the repair. The loss includes downtime, so what is worth paying for is fast replacement, not repair.
- Genuinely wider cover. If your own history says your devices die by being dropped, accidental-damage cover insures the right failure mode; the included warranty never did.
- Near-zero incremental cost. Cover bundled at no extra charge, or protection you already hold through a payment method or household policy.
The mirror image is just as clear. A plan that mostly duplicates the manufacturer's term, carries an excess close to the repair cost, or excludes the way the item will realistically fail has failed on its own terms — before any argument about averages.
Deciding it in a minute at the counter
- What does the included warranty already cover, and for how long? Subtract that from the plan's term.
- What is excluded — specifically accidental damage, wear and batteries?
- What is the excess, and any cap per claim or over the contract?
- What does a claim produce — repair, replacement, refurbished unit or credit — and how long does it take?
- Could you replace this tomorrow without pain? That is the buffer question, and it outranks the other four.
- If you decline, self-insure properly. Set the money aside rather than merely not spending it.
Step six is the one people skip, and it is what makes declining a strategy rather than a gamble: across a lifetime of purchases you keep the margin the insurer would have taken, and meet the occasional loss from the fund built for it.
FAQ
Is a phone plan different from an appliance plan?
The failure modes differ, so the answer does. Phones are dropped and soaked; appliances tend to fail mechanically once the defect window has closed. Accidental damage is the coverage most likely to be worth paying for on a phone, and the one most often missing from a basic plan.
Isn't peace of mind worth paying for?
Sometimes, and it is a legitimate thing to buy. Just check the policy delivers it. A plan with a meaningful excess, a slow claims process and a long exclusion list can leave the worry exactly where it was.
I bought a plan and never claimed. Was that a mistake?
Not necessarily. Judging a decision by how it turned out — what the former professional poker player Annie Duke calls "resulting" — mistakes luck for judgment. A plan you did not need can still have been sound if the loss would have been unaffordable at the time.
How do I self-insure without spending the money on something else?
Keep it somewhere separate, add to it every time you decline a plan, and draw on it only for the kind of loss it was created for. It has to be at least as large as the single biggest loss you have chosen not to insure.
Extended warranties are neither a trick nor usually a bargain. They are insurance sold in the least favourable place to buy it, and the decision gets simple once you stop asking whether the item might fail and start asking what its failure would cost you. Insure what would take you out of the game; carry the rest yourself, deliberately, with the money set aside. For the full entries on expected value, loss aversion and risk of ruin, visit Build Mind.
