In 2018 Zagat surveyed 13,000 restaurant guests. The number one complaint was noise - 24% of all grievances. Not food, not service, not prices. Noise.

Most restaurateurs respond to this the same way: "Well, we have a lively room." And carry on changing the menu.

The pattern repeats itself: a restaurant invests in the kitchen, the team, the interior - and guests stop coming back. The reason nobody checks first: restaurant acoustics. How the room sounds on a Friday at full capacity.

Why restaurant acoustics don't show up in reviews - and still drive guests away

A guest won't write "your RT60 is three seconds." They'll write "we couldn't hold a proper conversation" - or simply give four stars instead of five and leave for a competitor.

Researchers at the University of Nebraska showed that acoustic complaints in online reviews drag down a restaurant's overall rating by a full star - even when the food and service are excellent. Action on Hearing Loss found that 91% of guests said they would not return to a noisy restaurant. Eight out of ten had left earlier than planned.

The guest doesn't come back. And assumes the problem was the kitchen.

The restaurateur thinks the same - and changes the chef.

How noise changes taste: physiology, not atmosphere

This is the least obvious and the most important point.

Charles Spence is Professor of Experimental Psychology at the University of Oxford. His specialty: how different sensory signals interact in the human brain. His conclusion: taste is 50-75% shaped not by the chemistry in the mouth, but by the context of the space. Light, smell, touch - and sound.

When ambient noise exceeds 70 dB, taste perception deteriorates. This was confirmed by EEG and skin-conductance measurements at the University of Copenhagen (Frontiers in Neuroscience, 2022): above 70 dB, cognitive load rises and decision quality falls.

This is not about atmosphere. It is about physiology. A restaurant with poor acoustics literally prevents guests from judging the food fairly. The chef cooks excellent dishes - but the guest at the table is experiencing something else, acoustically altered.

The direct consequence for ordering follows from this. Biswas et al. (Journal of the Academy of Marketing Science, 2019): at high noise levels guests make impulsive decisions - they take the first thing they see on the menu. At low noise levels they read carefully, choose deliberately, and spend more. Noise switches the decision mode from "considered" to "urgent."

The Lombard Effect: why the problem compounds on its own

The Lombard Effect - a mechanism identified in 1909: as background noise rises, people instinctively raise their voices. By 0.3-0.6 dB for every extra dB of background. Automatically, without intention.

The space generates noise on its own - people only amplify it.

That is precisely why "asking guests to speak more quietly" does not work. And neither does changing the music. This is a structural problem: a room where sound reflects rather than is absorbed. Only the acoustics of the space can fix it.

Bottalico et al. (Nature/Scientific Reports, 2022): in rooms with a reverberation time below 0.7 seconds the Lombard Effect is significantly weaker. The feedback loop either never starts, or is damped at an early stage.

RT60: the one number to know before a project begins

RT60 is the reverberation time - how long sound lives in a room after the source stops.

Concrete, glass, open ceilings, timber floors - contemporary aesthetics and an acoustic problem by default. RT60 can be measured with a phone in 10-15 minutes (apps such as Room EQ Wizard). This is the baseline before any conversation with an architect.

Steffens, Wilczek, and Weinzierl (Frontiers in Built Environment, 2021) studied 12 restaurants in Berlin: RT60 affects the perceived quality of a restaurant no less than the absolute noise level. A guest doesn't hear "three seconds of reverberation" - they hear "some kind of oppressive drone" and leave early.

The first thing to address when RT60 is high: the ceiling. According to Resonics, treating 50-60% of ceiling area with acoustic materials delivers the greatest effect per unit of budget. Sound reflects downward from above - the ceiling is the primary reflecting plane. The right suspended panels work both acoustically and visually.

What happens to the order at 78 dB in the room

Cornell HRI (Wansink): when noise exceeds 78 dB, the number of items ordered drops by 18%. At 82 dB, staff error rates when taking orders rise by 34%.

Let's translate that into money. 60 covers, average spend €35, Friday evening. A drop of 18% in items ordered equals roughly €6 per table. At full occupancy that is around €360 the restaurant did not earn. Not because of the kitchen. Because noise was making decisions for the guests. Every Friday.

A separate note on music. Behavioral Sciences, 2024 - a field experiment in a real restaurant: slow tempo, an average of 80 minutes at the table; fast tempo, 57 minutes. A difference of 23 minutes. That is turnover management through the sound environment - an operational tool that either gets used intentionally or operates by accident.

Acoustics and concept: why it is designed in, not added on

Almost every question about restaurant acoustics comes up after opening. "It's somehow too noisy - how do we fix it?" is the standard situation. It can be fixed after opening - but that costs more, and it looks like solving a problem. The better approach is to build it into the project from the start: then it looks like design.

What to put in the brief: "Target RT60 - 0.7 seconds at full occupancy." That is a conversation an architect can work with - specific and measurable, unlike "we'd like good acoustics."

A few signals worth checking:

• Measure the RT60 in your room. If it is above 1.2-1.5 seconds, there is a problem.
• On a Friday evening at full capacity: try making a phone call. If it is difficult, the noise level is already above 78-80 dB.
• In your reviews: don't look for the word "noise" - look for phrases like "couldn't hold a proper conversation," "felt uncomfortable," "not quite our kind of place." Those are traces of an acoustic problem.
• Do guests at the table regularly lean forward? That is not a posture choice - it is an attempt to make out what someone is saying.

The goal is not to make the room quiet. A quiet restaurant is also a problem: emptiness, every conversation audible to the neighbours. The goal is controlled acoustics. Fine dining and a busy brasserie with an open kitchen should sound different - but both should sound that way intentionally, not by default.

In 2025, rd+d Magazine named "acoustic awareness" the number-one trend in restaurant design - ahead of biophilia, AI, and open kitchens. The market has arrived at something simple: guests pay for a good evening. A good evening includes how the room sounds.

A good room goes unnoticed. Guests don't say "their acoustics are excellent" - they say "it's good there, let's go back." Bad acoustics go unmentioned too. People simply don't return.

More material on evidence-based design is available in the Insights section, including on restaurant lighting. Completed projects are in the portfolio.