Skip to content
Launch offer — everything free, no limits. Ends in 10d 08:56:49
FragranceIQ

How our buy/wait verdicts actually work

No black box: exactly how a price becomes a verdict, and why every verdict comes with its reasons attached.

FragranceIQ Editorial · 4 min read

Every fragrance on FragranceIQ carries one of four verdicts: Good Deal, Buy Now, Wait, or Overpriced. These are not vibes — each verdict is produced by a fixed set of rules comparing today’s best price against that fragrance’s own history, and every verdict lists the exact reasons it fired.

We built it this way on purpose. Price-advice tools that cannot explain themselves are asking for blind trust, and blind trust is what got fragrance buyers into the RRP mess in the first place.

The signals we look at

The engine weighs a handful of signals for each bottle and size: the current lowest price across the retailers we track, the average market price for that same bottle, how deep today’s discount is compared with the typical discount over the last 90 days, how volatile the price has been, and how the current price compares with the lowest price seen in the recent window.

A discount only matters relative to a product’s own habits. Twenty percent off a fragrance that is discounted every week is background noise; the same twenty percent off a product that almost never moves is a genuine event. The engine treats those two situations very differently — which is exactly what an experienced deal-hunter does by instinct.

What each verdict means

Good Deal means today’s price is meaningfully below the product’s established market level — historically a strong buying moment. Buy Now is a step stronger: the price sits at or near the bottom of its recent range and prices like it have not lasted long historically. Wait means the current price is unremarkable against history and patience has usually been rewarded. Overpriced flags a price sitting clearly above the market level — often a single retailer out of step.

Confidence matters too: every verdict carries a confidence score reflecting how much history stands behind it. A verdict resting on two weeks of data tells you it is resting on two weeks of data.

What we do not do

We do not accept payment for verdicts, we do not rank retailers by commission, and we never let a business relationship touch the engine. The rules live in one configuration file, they apply identically to every product, and when the data is thin we say so rather than bluffing.