The Scent File Method
How we score a match against the original, and what stands behind each score.
Curated by the Scent File EditorWhat we score
A scored match carries two numbers from 1 to 10. We never average them into one, because that would hide the trade-off that matters most.
Accuracy: how closely the match smells like the original, from the first spray to the dry-down.
Longevity: how long the match lasts on skin in normal wear.
A very close match that fades in two hours is a different purchase from a looser one that lasts all day. We show both numbers so you can choose.
Where the scores come from
From people who have compared the two in public: mostly owners writing on Fragrantica and Reddit, and some reviewers. We read those reports and record the score, the page it came from and the month we last checked it.
Of the 494 matches scored from a source you can open, 374 cite a Fragrantica page, 64 a Reddit thread and 56 another forum, review site or video. For 328 of them we also keep the owner's own words.
A seller's own page is never counted as a source.
How a score is backed
Not every score has the same weight of evidence behind it. Cards across the site carry one of these tags, so you can tell which kind you are reading.
Three or more owner mentions. The tag carries the number.
Armaf Club de Nuit Blue Iconic, a match for Chanel Bleu de Chanel EDP
One or two owner mentions, or a source where we have not counted them. Real evidence, but thin: read it as a direction, not a verdict.
Armaf Club de Nuit Intense Man Parfum, a match for Creed Aventus
Scored, but we cannot yet point you to a source you can open and check. The bars are hatched and the score is called provisional wherever it appears.
Zara Gardenia, a match for Gucci Flora Gorgeous Gardenia
Listed, not scored. This includes every match whose only backing is the seller saying it is their version of the original: we name the pairing and show no number.
62% of the 801 matches we list carry a score with a source you can check. This page is counted from the data each time the site is built.
56 matches are scored for accuracy only: owners compared the smell, but nobody has reported how long it lasts. Those say "not reported yet" where the longevity number would be.
How the top match is chosen
Accuracy first. A tie goes to the match that lasts longer, then to the stronger projection, then to the lower price.
We call the top match The Scent File Pick only when its score has a source you can check. Otherwise it is the “highest-scored match”, or the “first listed match” when nothing is scored.
Who curates this
The Scent File is written and maintained under one byline, the Scent File Editor, and it is independent: not a brand, a dupe house or a retailer.
We don't claim to have worn all 801 matches ourselves. Nobody honestly could. What we do instead is show our work: the source, the owner's words where we have them, and the month we last read it. Where we have none of that, we say so and show no number. Spotted a score that looks wrong? Tell us.
How we stay independent
The Scent File earns a commission when you buy through our links to retailers. That is how the site is funded.
The commission has no effect on rankings. We do not rank matches by payout, and we do not accept payment for inclusion or position. A match that pays us nothing ranks above one that pays, if it scores higher.
One limit you should know about: since October 2026 we add a new match only if it is sold on Amazon, where our links earn a commission. That decides which matches get listed. It does not move a score or a rank.
What we don't do
No brand deals. No sponsored placements. No paid rankings. No single combined score that hides the trade-off between accuracy and longevity. No scores taken from sellers.