Three words get used interchangeably by people who should know better, including sellers. They mean different things, and the difference decides whether you annoy a room.
The definitions
Longevity is how long it stays detectable on skin. Measured in hours.
Projection is how far it travels from you right now. The radius of the cloud.
Sillage is the trail you leave behind, from the French for a boat's wake. What somebody notices after you have walked past.
They are independent, and a few familiar fragrances demonstrate it better than any explanation. Terre d'Hermès has good longevity and deliberately small projection: it lasts and sits close. Angel has enormous projection and sillage from the first minute. Cool Water projects hard early and fades faster than its reputation. Three fragrances, three completely different shapes, and none of them is better than the others.
What you want depends on the situation, and our guide to wearing fragrance covers that. What you almost never want is the maximum of all three: that is exactly what the marketing and most video reviews push.
How the industry actually measures this
This matters because it tells you how soft the numbers in reviews are.
The industry standard is headspace gas chromatography, often paired with mass spectrometry. A sample is placed in a chamber and the air above it is sampled repeatedly at set intervals, commonly 2, 6 and 24 hours, so the instrument can report which molecules are still present and at what concentration.
That is a machine reporting molecules in air. It is not a person reporting a smell, and the two answer different questions. When a reviewer says "eight hours," they mean they could still detect it, on their nose, on their skin, that day. Our guide to how smell actually works explains why that number is not transferable to you.
What actually drives longevity
The materials. Our notes, accords and the pyramid guide explained that small light molecules evaporate fast and large heavy ones do not. A fragrance built on substantial base materials, musks, resins, woods, heavy synthetics, lasts. One built on citrus and light florals does not, however much you pay. Our materials guide's citrus entry said this plainly: chemistry, not quality, and money does not fix it.
A clear pair: CK One and Shalimar. One is built on citrus and light musk, the other on benzoin, vanilla and resin. The difference in how long they last is a difference in molecular weight, not in care or budget.
The concentration, with the caveats in our guide to reading a fragrance description. More material in the alcohol generally lasts longer, but the categories have no legal definition and the ranges overlap, so a base-heavy eau de toilette will outlast a bright eau de parfum. Eau Sauvage is an EDT that outlasts plenty of modern EDPs, because Roudnitska built it on materials that stay.
The physics, briefly, because it is better than the folklore
"Heavy molecules last longer" is roughly right and it is not the whole model, and the real one explains things the folklore cannot.
Perfumery classifies notes by vapour pressure, which is a measure of how readily a substance leaves a liquid for the air. One fragrance patent gives the working bands: heart notes sit between 0.0008 and 0.08 Torr at 22°C, and base notes below 0.0008 Torr. Limonene, the citrus molecule, sits far above both. Vanillin sits right at the bottom: that is why it works as a fixative.
Then a proper model of what happens on skin. Kasting and Saiyasombati, in the International Journal of Cosmetic Science in 2001, built one from physical properties rather than assumption. It predicts evaporation from vapour pressure, molecular weight, the octanol/water partition coefficient and water solubility, using a modified Henry's Law with a boundary layer, and it separately models how much of the material is absorbed into the skin rather than lost to the air.
Volatility is not the only lever. How much a molecule likes fat matters too. A material that partitions into skin lipids stays put even if it is fairly volatile, which is what a fixative is doing and why the moisturised-skin trick works on physics rather than on hope.
Air velocity is in the model. The boundary layer of still air against your skin is what the molecules have to cross, and moving air strips it away. Which means a windy walk, a car vent aimed at you, a fan at your desk, or a gym all cost you fragrance in a way that has nothing to do with the bottle. Nobody says this and it is in the equations.
The practical version: if a fragrance seems to vanish, ask what the air was doing.
Skin chemistry: real, and smaller than claimed
Everyone in fragrance says skin chemistry changes everything. The truthful version is narrower.
What is well supported: Sebum. Oily skin holds fragrance longer, because oil traps aromatic molecules and releases them slowly. Trade figures put oily skin at six to eight hours or more against two to three on dry skin.
Temperature. Warm skin pushes more material into the air, so the same fragrance projects harder in July than in January, and harder on someone who runs warm.
Hydration. A moisturised surface holds fragrance better than a dry one. This is why the unscented-lotion-first trick works, and it is the least controversial performance tip there is.
What is overstated: the claim that your personal chemistry transforms a fragrance into something different is much weaker than its popularity suggests. A 2009 study in the Flavour and Fragrance Journal by Patel and colleagues found skin-mediated chemical transformation of perfume was low on clean, dry skin under normal conditions. The acid mantle matters, but it is not the dominant factor.
So the fair position is that your skin changes how much of a fragrance reaches the air and how long it stays there by a considerable amount, while changing what the fragrance actually is by rather less than the folklore claims, and the gap between those two effects is where most of the confusion lives.
Where a fragrance really does smell different on two people, the larger part of the explanation is in our guide to how smell actually works: your two noses are not the same instrument. Specific anosmia will do more to change a perceived fragrance than anybody's pH. Roughly one person in five is mildly anosmic to certain isomers of Iso E Super, which is in Terre d'Hermès among a great many others.
Why a fragrance seems to disappear
Usually it has not. Our guide to how smell actually works covered olfactory adaptation: your nose stops reporting a constant smell within minutes. You are the one person who cannot reliably tell whether your fragrance is still working.
Do not judge longevity on yourself. Ask someone else, or smell your sleeve or a scarf rather than your wrist, since fabric holds fragrance longer and your nose is less adapted to it.
The reapplication urge is almost always adaptation, not fade. This is the most common cause of over-application, and the people around you are getting far more than you think.
Fabric lies in the other direction: a jumper can hold a fragrance for days, but that is not longevity on skin.
Winter, which is most of the year here
Two effects pull in opposite directions and most advice only mentions one.
Cold slows evaporation. Less energy in the molecules, less material crossing into the air, so a fragrance sits closer to you and in principle sits there longer. That is the half everyone repeats.
Dry air undoes it. Heated indoor air in a Canadian winter routinely runs below 30% relative humidity. Dry skin holds less, and the trade consensus is that fragrance evaporates faster off a dry surface than a hydrated one.
So a January morning in Regina gives you a fragrance that projects less and may not last any longer, and people conclude their bottle has been reformulated. It has not. The room is dry and it is minus thirty outside.
What to do about it, and only the first is well founded. Moisturise, which addresses the actual mechanism. Wear the heavier, more base-driven things in winter: that is the received wisdom, and it is at least consistent with the physics above. And do not judge a new fragrance for the first time in February, because you will be testing the weather rather than the bottle.
We are labelling the humidity half of this as trade consensus rather than measured fact. Nobody has published the experiment we would want: the same fragrance on the same person at 20% and 60% humidity with gas chromatography at both ends.
Making a fragrance last, without buying a different one
In rough order of how much difference they make, starting with the single biggest improvement available: apply to moisturised skin, using an unscented lotion first.
Spray onto skin, not into the air. Walking through a mist wastes most of it.
Pick your spots for the outcome you want: chest and neck project toward other people, wrists and inner elbows sit closer, and behind the ears is warm, which raises projection.
Do not rub your wrists together. Our perfume myths guide explains why the usual reason given for this is wrong, and why the advice survives anyway.
Clothing holds longer than skin, but some materials stain and a fragrance on a daily coat will build up.
Store it properly. Our storage and care guide has the numbers: for every 10°C rise, reaction rates roughly double, and a bottle kept in a steamy bathroom is losing performance you paid for.
A word about the performance obsession
Our guide to the fragrance industry right now showed that video culture pushed the market toward loud and long-lasting, because those are the only qualities a video can demonstrate.
The result is people evaluating fragrance on the two axes with least to do with whether it is any good. Bandit is not improved by lasting fourteen hours. Diorissimo is one of the great achievements in our fragrance canon and it is quiet and does not last especially long, because a lily of the valley built from hydroxycitronellal was never going to.
Performance is a specification. Whether you want to smell it every day for two years is the actual question.
How to test performance properly
- One fragrance, one arm, two sprays. Never two at once; you cannot separate them at hour four.
- Note the time. Then check at 1, 4 and 8 hours, which mirrors the industry intervals above.
- Smell your sleeve, not your wrist, at each check. Your wrist has been under your nose all day.
- Ask one person at hour four whether they can smell anything and from how far. That is your projection number and it is the only one you cannot fake.
- Repeat on a different day in different weather. Heat changes everything, and one reading is an anecdote.
Do that once for a fragrance you own and you will trust your own data over any review.
Part of Nose to Bottle, our complete fragrance manual. Read it in order here.