In 2008 Starbucks pulled hot sandwiches from its coffee shops: melted cheese and warmed meat were overpowering the aroma of coffee, and coffee was the brand's main anchor. That was a first-layer scent decision. Most restaurants start with the third - they buy a diffuser with vanilla or lavender, set it by the door, and consider the subject closed. The gap between these two approaches is huge: one manages perception, the other just fakes it.

Smell is the fastest sensory channel

Every way we have of sensing the world runs a route through several stations. Sight: the retina, the optic nerve, the thalamus, the primary visual cortex. Hearing: the auditory nerve, the medial geniculate body of the thalamus. Taste: the facial and glossopharyngeal nerves, the medulla, the thalamus again. The thalamus sits in the middle of all of them - a dispatcher that processes the signal and decides what to pass on, and with what priority.

Smell has no such dispatcher. Odor molecules go straight to the olfactory bulb, and from there to the amygdala and the hippocampus. That is, directly into the zones of emotion, fear, pleasure, and long-term memory, bypassing analysis.

From this come two things that make scent a special tool.

The first is speed. The olfactory response happens before conscious evaluation kicks in. While still on the threshold, the guest is forming an emotional reaction and has already decided, on a subconscious level, how long they will stay and whether it's worth lingering at all.

The second is the link to taste. Charles Spence of Oxford, one of the biggest names in crossmodal perception (that's when one sense bleeds into another), shows that most of what a person perceives as "the taste of a dish" is actually smell, not what they feel on the tongue (Spence, Flavour, 2015).

You can see it with AirUp bottles: you pour in plain water, click a scent pod on top - and you drink water, but it tastes to you like something close to juice. There's zero sugar and zero flavoring in the liquid; smell assembles the taste. That same effect in pure form.

And it's not only about smell. The same crossmodal principle works with color: the color of the tableware and the walls shifts perceived sweetness and freshness by roughly 15% (the mechanism is broken down in color psychology). It's just that smell has the most direct and most powerful channel to taste.

And let me say it up front: this is not about the case where the room plainly stinks - that's obvious and it isn't fixed with marketing. This is about the subtle background a guest won't put into words, but which mixes into every dish and settles into the check.

The guest walks in. They haven't seen the room yet, haven't picked up the menu, haven't spoken to a server. The attitude toward the place is already there. By the time they open the menu, the verdict on the venue has already formed. You can correct it with other tools, but with a handicap.

That's exactly why scent is not one of the elements of atmosphere. It comes first.

One scent - plus 20% spend: data from a real restaurant

A pizzeria in Brittany, 88 guests, several evenings in a row. One venue, the same dishes, the same staff, the same prices. The only thing that changed: on some evenings the researchers diffused lavender into the room, on others a lemon scent, on others nothing.

The lavender evenings produced 20% more spend on food and drink compared with the control evenings. The lemon scent - a neutral result. (Gueguen & Petr, International Journal of Hospitality Management, 2006.)

This is not a lab experiment. Real guests, a real restaurant, real money.

And +20% off a table isn't an abstraction: time at the table converts into money directly. How much each extra minute is worth is broken down in restaurant zoning.

Why did lavender work and lemon didn't? Roschk and Hosseinpour analyzed 671 effects from scent experiments in retail and service spaces and gave the answer: congruence - the match between the scent and the venue's concept and format - determines whether the scent works at all. (Journal of Marketing, 2020.)

Lavender created the atmosphere of a calm, unhurried meal. The pizzeria in Brittany operated in exactly that format - a leisurely dinner, a table held for an hour or an hour and a half, a second glass of wine as the norm. Lavender matched the concept and reinforced the behavior the venue was already encouraging.

Lemon is a different story. Citrus scents are associated with alertness and a fast pace. They work in coffee shops and places with quick turnover. In a calm pizzeria, where the goal is to keep the guest longer, they don't.

A mismatch between scent and concept can even produce a negative result. A cedar scent in a sushi bar, heavy oriental spices in a minimalist Scandinavian cafe - the guest feels a cognitive dissonance. They don't think "there's a strange smell here." They think "something's off" - and that attitude carries over onto their judgment of the food, the service, the bill.

The question "what scent would suit us" is the second one. The first: what should the guest do in this zone - stay longer and order dessert, or, on the contrary, free up the table faster? The answer to that determines the direction of the scent. Not the owner's personal preferences.

Scent hints not only at how long the guest will stay, but at what they'll order

Biswas and Szocs set out to examine a different question. Not "does scent work on behavior," but how exactly it changes what people choose. (Journal of Marketing Research, 2019.)

They ran experiments with one and the same smell - the scent of freshly baked cookies. The variable was the duration of exposure.

Thirty seconds - participants chose more indulgent dishes. Two hours of the same scent - the choice shifted toward healthier eating: salads, lighter dishes, fewer desserts.

The mechanism is neural compensation. An olfactory stimulus activates the brain's reward system. With short exposure the system is "warmed up" and wants more - the person reaches for the rich and the fatty. With prolonged exposure the system gets enough stimulation through the nose, and real consumption becomes less appealing. The nose gets full before the body does.

For a restaurant this changes the job itself. The kitchen aroma seeping into the room is not background. It's already the start of the menu. It builds an appetite for a certain type of dish, or kills it, before the server has even reached the table.

The smell of fried onions and hot oil sets one vector: hearty, heavy food. In a burger joint or a solid bistro that works. In a restaurant built around light Mediterranean cooking, the smell from the kitchen creates a contradiction: the guest sees carpaccio and fish, but their sense of smell has already tuned them to something else. That friction affects both the choice and the pleasure taken in it.

A restaurant with an open kitchen gets all of this at full blast. The guest is in direct contact with the production smell all evening. If the kitchen is cooking something appetizing and congruent, it's live advertising - the appetite builds. If it's the smell of the deep fryer in a room with a check starting at 3000 rubles per person, neither the service nor the interior will compensate. Smell here is one of the channels that, with an open kitchen, are accounted for before the build (it also adds +7-13 dB to the room): reworking the extraction in an operating venue costs incomparably more than the figure on paper.

Three layers of scent: why the diffuser by the door doesn't solve the problem

When a restaurateur decides "we need scent marketing," they usually think about the third layer. They pick a smell they like: vanilla, lavender, cedar, citrus. They order a diffuser. And they wait for a result.

The third layer is finishing work. It only makes sense if the first two are already done.

Layer 1: neutralizing the negative. The most important work - and the least obvious, because it's invisible. A restaurant has several sources of negative smell. The kitchen: grease, overheated oil, onions, spices. The restroom: even in a clean venue, ventilation does its part. Upholstered furniture: sofas and chairs accumulate smells over the years - food, clothing, perfume. Ventilation: a smell of damp or dust that no one has noticed for ages. The waste area, if it's near the entrance or the ventilation pulls from it.

If these sources are active, a vanilla diffuser doesn't create atmosphere, it creates conflict. The guest feels something pleasant and something uneasy at the same time. The limbic system is evolutionarily wired to react to alarming signals more strongly than to pleasant ones - the unpleasant outweighs. The result is worse than if there had been no diffuser at all.

Layer 2: a neutral base. Clean air, with no smell, as a result in its own right. It's achieved through ventilation, regular dry-cleaning of upholstered furniture, properly working kitchen extraction, control of the restroom area. Restaurants that reach this level often already win: "it just felt good there, I can't explain it" - often that's simply the absence of anything extra. A neutral base doesn't need a diffuser. It just needs discipline in the daily operation.

Layer 3: an atmospheric scent. A congruent smell that reinforces the concept, sets the pace, creates an association with the place. This is where diffusers work, scented candles in certain zones, the kitchen aroma as part of the experience - not as a problem to hide. At this level scent becomes part of the venue's identity.

Starbucks in 2008 made a first-layer decision: it removed the source of the conflicting smell. A diffuser on top of frying sandwiches would have produced nothing - it would have added one more layer to the ones already in conflict. The order: first remove the bad, then build the neutral, then add the good. In exactly that order.

Why the owner has no access to the information every incoming guest has

Olfactory adaptation is a physiological mechanism in which the receptors lower their sensitivity to a constant background stimulus. It happens within 15-20 minutes of continuous exposure. An evolutionary mechanism: if a smell is constant and not increasing, it carries no new information about a threat, so the brain stops spending resources on it.

The practical consequence. The cook arrives at 9 a.m. By 9:20 they've adapted to the smell of the kitchen. The server arrives at 11. By 11:20 - to the smell of the room. The manager comes in every day. The owner, who's there every day, has over years of work adapted to every smell in the venue.

The guest walks in at 7:30 p.m. The first 20-30 seconds are the only moment when they feel all of it: the diffuser, the kitchen, the restroom, the upholstered furniture. It's in those seconds that the attitude forms which then influences the judgment of the food, the feel of the service, the decision to come back.

This information is available to every new guest. And almost never to those who work inside every day. Not because they're inattentive. Because of physiology.

The only way to get this data is to bring in someone with a cold nose and let them judge the venue as an incoming guest.

A scent audit: come in from the street after at least 2-3 hours away. Not through the kitchen, not through the service entrance - through the main entrance, like a guest. Stop for the first 30 seconds and record everything you can sense.

The second tool: a person who hasn't been in the venue that day. Don't ask for their opinion - ask for their first reflex: "what did you feel on the way in?" One answer will give you more than a month of observing from the inside.

At handover we check the smell as seriously as the light and the acoustics: with a cold nose at every stage, and where we can measure the air on site, we look at whether the dips in ventilation line up with the dips at the till in the same hours. Your own nose adapts; the chart doesn't.

Tuning scent is no simpler than working with light, and underrating it is just as costly. How light steers guest behavior and the average check is broken down in restaurant lighting.

The conversation about smell doesn't start with "what scent will we like." And not with choosing a diffuser.

It starts with a question: what does the guest feel 20 seconds after walking in? Do we have an answer - or do we not know, because we stopped feeling it long ago?

The first layer matters more than the third, the base more than the diffuser. And the data for the right decision walks through your door every evening - with every new guest. All that's left is to start catching it.