Everyone scores the deal. Nobody scores the shop. Once we hold 90 days of prices we can say how long a retailer’s “was” price actually held before the next sale started.
If the former price is the actual, bona fide price at which the article was offered to the public on a regular basis for a reasonably substantial period of time, it provides a legitimate basis for the advertising of a price comparison.16 CFR §233.1 · FTC Guides Against Deceptive Pricing
The first score publishes 90 days after the first observation. Not before.
We report what a price did. We never state why a retailer did it. “This reference price held 4 days in 90” is measurable. “They deliberately misled you” is not, and we don’t write it.
Enforced as a lint rule in the copy pipelineEvery score links to the complete timestamped observations it was computed from, plus the formula. If you can’t reproduce our number from data we show you, we shouldn’t be publishing it.
Score and raw data on the same page, alwaysA retailer sees an adverse score before it publishes, and their response goes up beside it, unedited.
Notice period before first publicationWe sample once a day. A price that dipped for six hours is one we missed, so we say “observed”, never “lowest”, and gaps are drawn as gaps.
1 batch per day, disclosed on every chartThe formula carries a version and an effective date. Any past score can be recomputed exactly. We change the method in public or not at all.
Method frozen 26 July 2026, before any score existedWhen we get one wrong we say so, dated, in a log that stays up. We are building that log.
Correction log, permanent