How discount codes and coupons actually work online

Stop thinking of a discount code as a secret word and think of it as a rule someone at a retailer wrote down. When you type it into a checkout box, a record is looked up, its conditions are compared against your basket, and the total moves only if every one of them is met.

Those conditions are the whole story: an end date, sometimes one category only, full-price items only, first-time customers only, one use per account. None of it appears on the listing you found the code on, which is why two people can try the same characters an hour apart and get different results. You are not hunting for a code that works, but for one whose rules match the order you were already going to place.

Percentage off. Scales with the basket, which is why retailers like it and why it can pull you upward.

Fixed amount off. A flat reduction tied to a minimum spend, worth proportionally more on a small order.

Free shipping. Removes the delivery charge rather than the goods; often the biggest saving on a light order.

Account-bound codes. First-order, loyalty and student codes belong to one person: dependable, rarely public, and the reason a friend's code fails for you.

Almost every real code comes from the retailer itself: a welcome discount traded for an email address, a seasonal event published openly, a loyalty reward tied to one account, or a brand's own social feed. Codes also travel through partner and creator arrangements, and the further one gets from the shop that issued it, the likelier it is to arrive expired.

None of it is generosity. A code is a marketing instrument with a job, and the job explains the rules: a first-order code buys a customer, a cart-abandonment code recovers an order nearly won, a free-shipping threshold nudges the basket upward, a seasonal code clears stock. Ordinary commerce, disclosed in the terms — but built to change your behaviour, not simply to reward it.

Anyone who has pasted a dozen codes into a checkout box knows the ratio is poor, and the reasons are structural. Listings are mostly user-submitted or scraped and almost never re-tested, so dead codes stay visible; account-bound codes get reposted and cannot work for anyone else; and some strings are simply invented to make a page look stocked.

Honest signals exist. A code on the retailer's own page, in an email addressed to you, or inside your account is far likelier to be live, and published terms are a good sign, because a shop describing conditions is describing a code that exists. Even then, the only test that settles it is the total you are about to pay.

The most expensive trap is the most ordinary: spending more in order to save. A percentage code on a basket you padded out is not a discount, it is a larger purchase with a smaller sticker. The test is unforgiving — would you have bought this at full price? If not, the code saved you nothing.

The second is any page that asks for something valuable before it will show you a code. A card number, a bank login, an identity document, a one-time passcode: none is ever required to receive a retailer's discount, so treat the request as the end of the interaction — and read a browser extension's permissions the same way.

The third is quieter. Sale items are often excluded from promotions, so a code applied to full-price goods can cost more than the discounted version of the same thing, and stacking rarely rescues it, because most checkouts accept one code at a time.

None of this needs to become a hobby. Fill the basket with only what you came for and note the total — that is your baseline. Check your own mail and account first, then the retailer's own pages, then try what you have one code at a time, reading the total rather than the message.

If a code fails, read the error: below minimum, excluded items, new customers only, already discounted. None of those is fixed by trying harder. Then ask the only question that matters — is this basket still the one you wanted when you started?