The mechanism, not the conspiracy
How fake discounts work is duller than most people expect, and the dullness is exactly why the practice survives. Nobody sits in a room inventing lies about prices. What happens instead is that a retailer needs a number to measure a markdown against, nothing obliges that number to be a price anyone paid, and every price underneath it then looks like a saving. Permanently, automatically, and with arithmetic that is entirely correct.
That is the whole trick, and it has three common forms. Understanding them turns a sale badge from a signal into a claim you can check in about thirty seconds. The wider method for judging any price, including what a real saving looks like, is the pillar this sits under: is this a good deal.
Form one: the reference price nobody pays
Every discount is a subtraction, and the interesting half is what it subtracts from.
A retailer can use several things as that reference. A manufacturer's suggested price, set once at launch and often never revised. Their own standard price for the item, which is theirs to set at any level. Or simply the highest number the listing has ever displayed.
None of those has to be a price the product sold at in volume, and for an older product the gap can be enormous without anyone behaving badly. A suggested price from three years ago describes a market that no longer exists. The manufacturer has not updated it because nobody asks them to, and the retailer is not obliged to ignore it.
What makes it a fake discount rather than a stale number is the presentation. A $50 offer against a $200 reference shows a 75% badge. If the product has spent the last six months trading between $48 and $54, the saving against what it actually costs is close to zero. The badge is not lying about the subtraction. It is lying about what the subtraction means.
Form two: the sale that never ends
The second form needs no inflated reference at all. It just never stops.
If a listing carries a discount every time you look at it, across weeks and months, then the discounted price is the price. There is no moment at which you could buy it for the crossed-out figure, which means the sale is not an event you are catching. It is the permanent state of the page.
This one is the easiest to check and the most often dismissed, because it does not feel like a trick. Nobody is misstating anything. The product genuinely costs less than its reference, and it always will. What has been removed is the information a discount is supposed to carry, which is that now is different from usual.
Category matters here. In fashion and homeware, near-permanent markdown is close to standard practice and shoppers have largely adjusted. In electronics it is rarer and more informative, because prices there move for real reasons and a flat line under a permanent badge stands out.
Form three: the pre-sale rise
The third form is the only one that requires a deliberate sequence, and it is the one a chart makes obvious.
A listing drifts upward over two or three weeks. Then, at the start of a promotional period, it falls back to roughly where it began, with a large percentage attached to the fall. The discount is real against the raised price and meaningless against the price the product held a month earlier.
On a price history this looks like a hill rather than a cliff, and once you have seen one you will recognise the shape immediately. It clusters predictably in the weeks before major shopping events, which is why the first half of November is the best time to look at charts rather than badges.
Regulators in several countries have responded to this by requiring that an advertised reference be a price actually charged within a recent window. Where such a rule applies, the pre-sale rise becomes a compliance exercise rather than a prohibition: hold the higher price for the required period, then discount from it. The rule changes the timeline more than the practice.
Why the psychology does the heavy lifting
None of this would work if the crossed-out number were simply ignored, and it is not, including by people who know exactly what it is.
The first number you see sets the scale for every number after it. A jacket at $89 is judged against your sense of what jackets cost. The same jacket at $89 with $220 struck through is judged against $220, and it feels like a different purchase. The item has not changed; the frame has. That effect is robust, well documented and largely immune to knowing about it, which is why the defence has to be procedural rather than mental. The mechanism itself is covered in how anchor pricing shapes what a price feels like.
The procedural defence is to supply your own anchor before the page supplies one. Decide what you would pay before opening the listing, or look at the price history first and let that set the scale. Both work for the same reason: they put a real number in front of the invented one, so the invented one has something to be measured against rather than the other way round.
Urgency works on a separate lever. A countdown or a low-stock warning does not change what you think the price means. It changes how long you spend thinking, and thinking time is where a badge gets checked. A deal that is genuinely good survives ten minutes of checking. One that stops being good when you check was telling you something.
What none of this proves
It is worth being precise about the limits, because the honest version of this argument is more useful than the loud one.
An inflated reference is evidence about the reference. It is not evidence that a seller intended to mislead anyone. Suggested prices go unused for years. One product page can carry several configurations at different prices. A genuine long-run decline leaves the old list price behind without anyone deciding anything.
It also does not mean discounts are worthless. Clearance before a new model, genuine overstock and retailer-funded promotions all produce substantial, real reductions, and treating every badge as fake costs you those. The point of knowing the three forms is not to distrust everything. It is to be able to tell them apart quickly.
The check that settles it
Ignore the crossed-out figure entirely and ask one question: how does today's price compare with what this product has actually been selling for?
That needs the product's own tracked history rather than the listing. Look at the range over the last year, the band the price spends most of its time in, and where today sits against that band. A price clearly below the band is a real saving whatever the badge says. A price inside the band is the normal price with a percentage on it.
The opposite case is worth watching for as well, because it is where the money is. A product quietly at its lowest tracked price with no badge at all will be buried by any list sorted by discount, and it is frequently the best available purchase on the page.
To run that comparison on one specific listing rather than in the abstract, hand it to the fake discount checker.
