How to spot a fake discount
Inflated list prices, permanent sales and pre-event markups: what U.S. law requires of a "was" price, and how to check one yourself.
By Ahsan Khan, Founder. Published . 6-minute read.
Summary
A crossed-out price is not decoration. It is a factual claim that the item recently sold at that price, and U.S. advertising rules treat it as one. The claim goes wrong in a few predictable ways: an inflated list price nothing ever sold at, a sale that never ends, and a quiet markup before a big shopping event.
We test every "was" price against our own observations. It counts only if it was the prevailing price for at least 30% of the last 90 days; if it fails, we relabel it "Retailer's list price" and score the offer down. You can run a rougher version of that check yourself.
What a struck-through price is claiming
Every discount is two numbers: the price you pay and a reference price that makes it look like a saving. The first is verifiable at checkout; the second is a claim about the past, or about the wider market, and it is the half of the pair nobody checks. It can be honest in three ways, which are not interchangeable: the retailer's own former price, what other stores charge for the identical item, or the manufacturer's list price. Most deceptive pricing is a quiet substitution of one for another.
The six mechanics worth recognizing
- The fictitious list price. A number set high enough that the item has never sold there, so every day can be a sale day. The tell is a reference price that never moves while the selling price wanders underneath it.
- The permanent sale. A discount that resets weekly, with a countdown that has expired every Sunday for a year. If an item is continuously sold at the sale price, that is simply the price.
- The pre-event markup. The price rises quietly two or three weeks before a major event, falls back, and the difference is advertised as the event discount. Only price history exposes it, which is why our Black Friday window opens three weeks early.
- The anchor swap. "Compare at" pricing measured against a different retailer, a different configuration, or a bundle you did not want. The comparison is real; the thing compared is not the thing in your cart.
- The near-identical model. A derivative model number built for a promotional season, a notch below the version the reviews describe, with no history anywhere to check. This is why we key products on the manufacturer part number.
- The "up to" headline, where the largest figure belongs to one clearance item in the corner of the sale.
What U.S. law actually says
These rules bind sellers, not shoppers, and public enforcement is sparse. Read them as a description of what a defensible discount looks like.
| Rule | In plain language |
|---|---|
| FTC Guides Against Deceptive Pricing, 16 CFR 233.1 | A "was" price must be a bona fide price at which the item was openly and actively offered for a reasonably substantial period, in the recent regular course of business. A price set high in order to be crossed out is deceptive even if the item briefly carried it. |
| 16 CFR 233.2 and 233.3 | "Compare at" claims must use prices other sellers in the area actually charge for the identical item, and a suggested retail price may not be a reference when it is not what the item generally sells for. |
| California Business & Professions Code 17501 | A former price may not be advertised unless it was the prevailing market price within three months before the advertisement, or the advertisement says when it did prevail. This is the 90-day clock behind "was". |
The Guides are interpretive: they explain how the Commission reads the ban on deceptive practices in Section 5 of the FTC Act, and cases are brought under Section 5. California's provision is a statute, and because national retailers do not run separate California pricing, its three-month standard shapes careful behavior everywhere.
How we turn that rule into arithmetic
A retailer's "was" price counts only if it was the prevailing price for at least 30% of the last 90 days in our own observations. Clear that bar and we show it as a former price. Fail it and we relabel it "Retailer's list price" and mark the offer down, because authenticity is 35% of the FlexiScore, its largest component.
The 30% threshold is a compromise worth being honest about. Requiring a majority of the window would fail honest retailers, whose list price alternates with legitimate promotions. Requiring a single sighting would pass every markup.
Illustrative example. These figures are invented to show the arithmetic. They are not an offer, and they do not describe any real product.
- A hypothetical monitor, recorded at $400 on 62 of the last 90 days and at $340 on 28. Advertised today at $340, was $400.
- $400 held for about 69% of the window, which clears the threshold, so the former price is genuine and the $60 saving stands.
- Change one thing: the advertisement says was $520, which we never observed. Prevailing share is zero, so $520 becomes the retailer's list price and the claimed $180 saving does not survive. You pay $340 either way.
How to check a discount without us
- Search the manufacturer part number rather than the marketing name. If the promotional model appears nowhere else, the reference price cannot be verified at all.
- Check the identical item at two other retailers. A genuine cut spreads within days, because retailers watch each other. A cut at one seller, against an anchor nobody else uses, is the pattern to distrust.
- Look up the brand's own price on the brand's own site. If the retailer's reference sits far above it, that is not a former price; it is a number.
- Look at the item three or four weeks before a major event and write it down. A markup is invisible unless somebody recorded the before.
- Read the fine print for an end date, then come back after it. A sale that ends and immediately restarts on the same terms is the regular price in a costume.
- Screenshot the page with the date visible if you might price-match later.
For Amazon listings, independent trackers such as CamelCamelCamel and Keepa hold long histories. Elsewhere, the Internet Archive's Wayback Machine often has snapshots from before the event. Neither is complete; either is enough to catch a markup.
What a genuine cut looks like
Real discounts have causes: a successor has shipped, the model year has turned over, a competitor has been matched, or stock is being cleared as open-box. The price falls and stays down, and you can name the reason. A discount with no cause is the one to look at twice.
Questions shoppers ask
Is it illegal for a store to show a list price nobody pays?
Showing a manufacturer's suggested price is not unlawful in itself. Using it as a reference to imply a saving, when the item does not generally sell at that price, is what the FTC's guidance treats as deceptive. That is why we relabel it rather than hide it.
What if you have not tracked something long enough to test its discount?
Then we say so. Below 14 days of history or 30 observations, a product carries the "Early read" label, the verdict is explicitly not definitive, and the former-price test is reported as untested rather than as passed.
A retailer failed your test. Did they break the law?
No, and we do not say that anywhere on the site. Our test runs on our own observations, which may be incomplete, and uses one fixed threshold where the law asks a broader question. A failed test means we could not verify the claim, so we decline to repeat it.
Where do I report pricing I think is genuinely deceptive?
The FTC takes consumer reports at reportfraud.ftc.gov, and every state has an attorney general's consumer protection office. Bring dated screenshots of both prices: without the before, a complaint is hard to act on.
What this is built on
- FTC Guides Against Deceptive Pricing, 16 CFR Part 233
- California Business & Professions Code 17501
- config/score-weights.yaml — The window, the threshold and the fallback label on this page are read from it at build time.
Putting this to use
- How we score a deal — the published weights, the verdict thresholds, and the former-price test this guide refers to.
- How we make money — affiliate links, and the architectural reason a commission cannot move a verdict.
- All guides and explainers
Bylined and human-reviewed, not generated per product: 2% of machine-written sections on this site are sampled for review, and 100% of guides are. Editorial policy and corrections.