Do 'Dupes' Copy the Original's Notes? We Compared 34 Published Pyramids Side by Side

Everyone in this hobby has said some version of it: a dupe is the original’s note list, cheaper. The dupe subreddits argue the opposite: that notes mean nothing on a clone. Neither side has had the data, because settling it requires cited note pyramids for both halves of a pair, from the houses themselves.

We have that corpus. So we ran it: 34 clone→original pairs, every one of which had a published pyramid on both sides, compared note list to note list.

Neither side of the argument wins. A typical dupe lists about 33% of the notes the original’s own house publishes, and the range runs from every single one to not one.

How much do a dupe’s notes actually overlap the original’s?

Here is the whole set. The first number is the one worth reading: of the notes the original’s house publishes, how many does the dupe also list?

Fragrance	Said to imitate	Of the original’s notes, also listed	Overlap of the two full lists	Shared notes in same tier
Afnan Supremacy in Oud	Initio Parfums Privés Oud for Greatness	100% (6 of 6)	100%	n/a
Armaf Club de Nuit Intense Man	Creed Aventus	89% (8 of 9)	62%	75%
Afnan Supremacy Incense	Amouage Interlude Man	82% (9 of 11)	75%	100%
Afnan Supremacy Silver	Creed Aventus	67% (6 of 9)	50%	83%
Armaf Club de Nuit Imperiale	Parfums de Marly Delina Exclusif	67% (6 of 9)	43%	50%
Afnan Supremacy In Heaven	Creed Silver Mountain Water	62% (5 of 8)	50%	80%
Afnan 9 PM	Jean Paul Gaultier Ultra Male	60% (3 of 5)	25%	100%
Lattafa Ana Abiyedh	Xerjoff Erba Pura	57% (4 of 7)	44%	75%
Afnan Rare Carbon	Tom Ford Ombre Leather	50% (4 of 8)	27%	n/a
Armaf Club de Nuit Sillage	Creed Silver Mountain Water	50% (4 of 8)	24%	100%
Lattafa Give Me Gourmand Vanilla Freak	Maison Mataha Escapade Gourmande	50% (2 of 4)	33%	50%
Lattafa Her Confession	Les Liquides Imaginaires Blanche Bête	45% (5 of 11)	36%	100%
Afnan Supremacy Not Only Intense	Creed Aventus	44% (4 of 9)	29%	75%
Al Haramain Amber Oud Carbon	Creed Green Irish Tweed	42% (5 of 12)	29%	80%
Al Haramain Amber Oud Gold Edition	Xerjoff Erba Pura	43% (3 of 7)	23%	67%
Lattafa Khamrah	Kilian Angels’ Share	40% (2 of 5)	12%	50%
Al Haramain Amber Oud Rouge	Maison Francis Kurkdjian Baccarat Rouge 540	33% (1 of 3)	11%	n/a
Armaf Club de Nuit Private Key To My Love	Maison Francis Kurkdjian Baccarat Rouge 540	33% (1 of 3)	9%	n/a
Lattafa Yara	Christian Dior Poison Girl	33% (1 of 3)	10%	100%
Armaf Club de Nuit Private Key To My Life	Creed Green Irish Tweed	25% (3 of 12)	10%	100%
Armaf Club de Nuit Urban Man	Dior Sauvage	25% (1 of 4)	7%	n/a
Lattafa Badee Al Oud Sublime	Kayali Eden Juicy Apple 01	25% (3 of 12)	18%	67%
Maison Alhambra La Charmante	Parfums de Marly Cassili	20% (2 of 10)	11%	100%
Lattafa Asad	Dior Sauvage Elixir	20% (2 of 10)	11%	100%
Lattafa Give Me Gourmand Mallow Madness	Kilian Love, Don’t Be Shy	20% (1 of 5)	9%	0%
Al Haramain Amber Oud Exclusif Bleu	Parfums de Marly Layton	10% (1 of 10)	5%	100%
Afnan 9 AM	Maison Francis Kurkdjian Aqua Vitae Forte	10% (1 of 10)	5%	n/a
Armaf Club de Nuit Private Key To My Dreams	Tom Ford Bitter Peach	0% (0 of 7)	0%	n/a
Rasasi Hawas Ice	Paco Rabanne Invictus Platinum	0% (0 of 6)	0%	n/a
Lattafa Give Me Gourmand Choco Overdose	Akro Awake	0% (0 of 4)	0%	n/a
Lattafa Asad Zanzibar	Jean Paul Gaultier Le Beau Le Parfum	0% (0 of 3)	0%	n/a
Armaf Club de Nuit Private Key To My Soul	Initio Parfums Privés Side Effect	0% (0 of 6)	0%	n/a
Rasasi Hawas for Him	Paco Rabanne Invictus Aqua	0% (0 of 3)	0%	n/a
Lattafa Ana Abiyedh Rouge	Maison Francis Kurkdjian Baccarat Rouge 540	0% (0 of 3)	0%	n/a

Median: 33%. Eleven of thirty-four list half or more. Seven list nothing at all.

Read the two overlap columns together, because they answer different questions. The first asks how much of the original the dupe reproduces. The second asks how similar the two lists are overall, and it runs lower throughout for a mundane reason worth knowing: the clone houses publish longer note lists than the luxury houses do. Kurkdjian names three materials for Baccarat Rouge 540; its dupes name seven to nine. A dupe can list everything the original does and still look like a 30% match on the second measure.

“n/a” in the last column marks a pair where tier agreement cannot be computed. Either the original published no pyramid to compare positions against (ten pairs, which turns out to be its own finding) or the two share no notes at all, so there is no position to agree on.

Which pair is the closest match?

Armaf Club de Nuit Intense Man. Creed publishes nine notes for Aventus and Armaf lists eight of them, with 75% of the shared notes in the same tier of the pyramid. Bergamot and lemon up top, blackcurrant, pineapple, birch and jasmine through the middle, musk and patchouli underneath. That is not a coincidence, and it is not our inference. It is what both houses say in public.

Creed Aventus (Creed official)	Armaf Club de Nuit Intense Man (Armaf official)
Top: Calabrian bergamot, Sicilian lemon, blackcurrant leaf	Top: apple, bergamot, blackcurrant, lemon, pineapple
Heart: pineapple, pink pepper, jasmine	Heart: birch, jasmine, rose
Base: birch, patchouli, musk	Base: ambergris, musk, patchouli, vanilla

Close behind, and arguably more striking because the original is a niche house rather than a famous one: Afnan Supremacy Incense lists 9 of the 11 notes Amouage publishes for Interlude Man, and every single shared note sits in the same tier.

One pair goes further still: Afnan Supremacy in Oud and Initio Parfums Privés Oud for Greatness publish the same six materials and nothing else: lavender, nutmeg, saffron, oud, musk, patchouli. It is the only exact match in the set. Initio publishes its list flat, so there is no pyramid position to compare.

Creed Aventus is the one original where high resemblance is a pattern rather than a single case: its three dupes rank second, fourth and thirteenth of the 34, at 89%, 67% and 44%. Creed Silver Mountain Water and Xerjoff Erba Pura each carry two dupes clustered in the middle of the table.

And the furthest apart?

Baccarat Rouge 540, which is the result that should retire the assumption. Three different houses have releases widely called BR540 dupes. Kurkdjian publishes exactly three materials. None of the three dupes lists more than one of them.

Fragrance	Published notes	Of BR540’s 3 notes
Maison Francis Kurkdjian Baccarat Rouge 540	Hedione, Ambroxan™, Saffron	n/a
Lattafa Ana Abiyedh Rouge	bergamot, kumquat, nashi pear, caramel, geranium, ambergris, musk, oakmoss	none
Armaf Club de Nuit Private Key To My Love	cassis, peach, pear, amberwood, lily-of-the-valley, musk, patchouli, sandalwood, vanilla	1 (amberwood)
Al Haramain Amber Oud Rouge	fruity notes, floral notes, saffron, spices, amber, musk, oud	1 (saffron)

Lattafa’s list and Kurkdjian’s have not one material in common. Armaf’s single point of contact is amberwood, a trade name for the same ambroxan-family molecule Kurkdjian names, the one entry in the original that is a molecule rather than a note. Al Haramain’s is saffron.

Three houses, one reference, and on paper three unrelated compositions.

The same house can sit at both extremes

The sharpest illustration that the note list is not a reliable signal comes from a single original house. Initio Parfums Privés appears twice in this study:

• Oud for Greatness. Afnan’s Supremacy in Oud publishes exactly the same six materials. A perfect paper match.
• Side Effect. Armaf’s Private Key To My Soul publishes nothing in common with it at all. Zero of six.

Same original house, same clone tier, same kind of product page, opposite results. Whatever the note list is measuring, it is not “how hard did this house try to copy”.

Doesn’t that just mean they’re hiding the recipe?

It could, and published data cannot rule it out. But the provenance points somewhere less cynical. 31 of the 34 clone-side pyramids in this study come from the house’s own website: Armaf (8), Afnan (8), Lattafa (11, across its own and its US site), Rasasi (2) and Al Haramain (2). Only three came from an authorized retailer instead.

These houses publish their own note data, under their own names, for their own releases. That is what a fragrance house does. And the ones that resemble an original most closely are publishing that resemblance themselves: Armaf is the source for the eight-of-nine Aventus overlap, and Afnan for the nine-of-eleven Amouage one.

The other half of the same point: eight of the originals publish no pyramid at all, covering 10 of the 34 pairs. Baccarat Rouge 540 lists three materials with no top, heart or base. Dior Sauvage, Tom Ford Ombre Leather, Tom Ford Bitter Peach, MFK’s own Aqua Vitae Forte, Jean Paul Gaultier’s Le Beau Le Parfum and both Initio fragrances publish flat lists too. In this set the luxury and niche houses are frequently the less forthcoming party.

When a note is shared, does it stay in the same place?

Usually. Across the 21 pairs where the comparison is possible, meaning the original publishes a real top/heart/base pyramid and at least one note is shared, tier agreement averages 79%. Nine of those twenty-one are at 100%.

So shared material is not scattered. When a house lists a note the original also lists, it tends to list it in the same position, and nobody appears to be taking the original’s pyramid and shuffling it.

We report tier agreement only on those 21 pairs deliberately. A pair whose original published a flat list scores 0% by arithmetic, not by evidence. There is no tier there to disagree with. Averaging those in gives 49% and understates the result, which is exactly the sort of number that gets quoted without its denominator.

What this means if you are shopping

The note list is a weaker test than it looks, and the reason is the spread rather than the average. Two fragrances marketed the same way, at the same price, aimed at the same original, can list 89% of its notes or none of them. Nothing on the bottle or the product page tells you which of those you are holding.

So use the list for what it can support:

• A high share of the original’s notes and high tier agreement, as in Club de Nuit Intense Man against Aventus and Supremacy Incense against Interlude Man, is real evidence that two houses set out to build parallel compositions. Both of them said so in public.
• A low share is close to uninformative. It is not evidence of a different scent any more than a high share is proof of the same one; a composition can be rebuilt from different materials, and the published list is not the formula.

For everything in between, the accord families and your own nose are better instruments than the pyramid.

How we know this

Every number above is measured, and the measurement is narrow enough to state exactly.

The pairs. 34 clone→original pairs, identified by a human and published here in full so they can be challenged. They are not derived from a retailer’s “inspired by” field, and they are not our matcher’s output. Every house in this study sells under its own name and makes no such claim. All 34 scored. Two of them, Lattafa Yara and Al Haramain Amber Oud Carbon, sit on catalog rows whose names our automatic matching cannot fold onto the pair (one repeats the brand in the product name, the other carries a retailer suffix), so those two are pinned to an explicit product record by hand rather than matched by name. We did that rather than loosen the name matching, because a search for the Yara pair returns four products and two of them are different fragrances.

The two overlap measures. Share of the original’s notes is the count of notes on both lists divided by the number the original publishes. That is the directional question, and the one a reader checks by hand. Overlap of the two full lists is Jaccard similarity: the same shared count divided by every distinct note the two publish between them. The second is sensitive to list length, and because clone houses publish longer lists than the luxury houses, it reads systematically lower. Both are reported for all 34 pairs so neither can be quoted alone. Tier agreement is the share of shared notes appearing in the same tier on both sides, reported only for the 21 pairs where the original publishes tiers and the pair shares at least one note.

Canonicalization. Notes are compared as canonical material names, so spelling and synonym variants collapse: blackcurrant and cassis, agarwood oil and oud, amberwood and ambroxan all count as matches. That choice raises every overlap figure, which means no result here is an artifact of two houses spelling the same material differently.

The sources. Both sides are cited from a brand-official page or an authorized retailer, never a crowd-sourced database. On the clone side: 31 of 34 brand-official, 3 from an authorized retailer. Originals are brand-official throughout.

The limit that matters most. A published note list is a marketing document, not a gas chromatograph. It is often incomplete, sometimes poetic, and never audited. This study measures what these houses publish about their own fragrances. Nothing here establishes that any two of them do or do not smell alike, in either direction. A 0% row is not proof of a different scent, and the one 100% row is not proof of an identical one. What it does establish is that the note list, the single most-quoted piece of evidence in every dupe comparison online, is far too inconsistent to carry the weight put on it.