MelloJam Listen now

8 Restaurant Music Statistics Every Owner Should Know From a 282-Table Field Experiment

We read the full field experiment "How Does Background Music Affect Dining Duration, Tips and Bill Amounts in Restaurants?" The researchers observed real restaurant behavior across 282 tables and compared fast-tempo music, slow-tempo music, and the restaurant's regular playlist.

The result is useful because it is specific. Music tempo changed how long guests stayed, but it did not produce a statistically significant difference in bill amount. The study also found a controlled tip-amount effect for fast music. Those are three separate outcomes, and none of them should be turned into a universal promise that background music raises restaurant revenue.

Back to Restaurant statistics

Condition Tables Mean time in restaurant Mean bill amount Mean tip percentage Statistical result Setting Limitation
Fast music 99 57.29 minutes 197.31 ILS 14.57% Time was lower than slow and control, p < 0.01; controlled tip amount was higher than control, B = 2.219, p = 0.035; no significant bill effect One Italian restaurant in central Tel Aviv; evening service; parties of 1-2 adults A single restaurant and country; no customer questionnaire
Slow music 96 80.30 minutes 207.36 ILS 14.06% Time was higher than fast and control, p < 0.01; no significant bill effect Same setting A longer stay is not automatically more revenue when tables are scarce
Control playlist 87 69.22 minutes 208.79 ILS 13.65% Reference condition for the regressions; no tempo treatment The restaurant's regular Italian and Greek playlist Regular playlist mixed fast and slow songs, so it is not silence

The paper reports bill and tip amounts in Israeli shekels, or ILS. It notes that ILS 1 was about USD 0.27 at the time of the experiment, but raw currency comparisons should not be converted using today's exchange rate or treated as a current restaurant benchmark.

The experiment observed 282 tables, not 282 individual diners

The study analyzed 282 tables: 99 in the fast-music condition, 96 in the slow-music condition, and 87 in the control condition. The parties were limited to one or two adults to reduce peer effects, and the patrons did not know they were part of an experiment. The researchers collected time, bill, and tip information through field observation and the restaurant's payment system rather than relying on customer memory.

That denominator matters. The unit of analysis is the table, and bill and tip amounts are for the entire table. A table study can tell an owner how a music condition was associated with table behavior in that setting, but it cannot by itself estimate the effect on every diner, every restaurant format, or every daypart.

Slow music kept tables for 80.30 minutes, compared with 57.29 minutes for fast music

Mean time in the restaurant was 80.30 minutes under slow music, 57.29 minutes under fast music, and 69.22 minutes under the control playlist. The differences were statistically significant: slow music produced longer stays than fast music and control, while fast music produced shorter stays than both slow music and control, with the reported comparisons at p < 0.01.

The study reports that the slow-music stay was 40.2% longer than the fast-music stay, the control stay was 20.8% longer than the fast-music stay, and the slow-music stay was 16% longer than the control stay. These are duration effects, not revenue effects. They are most relevant when an owner is deciding whether a service period needs faster table turnover or a more relaxed experience.

Fast music shortened the average stay by 23.01 minutes versus slow music

The difference between the slow and fast means was 23.01 minutes per table. In a restaurant with a queue and strong demand, a shorter stay could create more opportunities to seat new parties. In a restaurant with spare capacity, the same shorter stay could simply mean less time for guests to order another course, another drink, or dessert.

The paper presents fast tempo as a possible tool for busy periods, but the experiment did not measure the number of waiting parties, unused tables, cleaning time, lost orders, or actual table-turnover revenue. Owners should therefore treat the 23.01-minute difference as an operational input and test its effect against their own demand and capacity constraints.

Music tempo did not produce a statistically significant difference in bill amount

The descriptive mean bill amounts were 197.31 ILS for fast music, 207.36 ILS for slow music, and 208.79 ILS for the control group. Despite those different sample means, the regression models found no significant effect of fast or slow tempo on bill amount. The paper reports that the result remained insignificant both when time spent was included and when it was removed from the model.

This is the article's most important guardrail. Slow music did not reliably make guests spend more in this field experiment, and fast music did not reliably make them spend less. Alcohol consumption and time spent were associated with bill amount in one model, but the tempo treatment itself was not. Do not turn the raw mean differences into a menu-sales benchmark.

Fast tempo raised the controlled tip amount, but the raw tip percentages were close

The raw mean tip percentages were 14.57% with fast music, 14.06% with slow music, and 13.65% in the control group. The study's regression model for tip amount found a positive fast-tempo coefficient of B = 2.219 relative to the control condition, with p = 0.035, after controlling for bill amount, time, alcohol consumption, and payment variables included in the model.

That result does not mean that fast music universally increases tips by 2.219%, or that the same effect will appear in another restaurant. The dependent variable in the model was tip amount, while the descriptive table also reports tip percentage. The difference is important: a larger bill can raise the tip amount even when the percentage does not move, and the study was not designed to establish a general tipping rule.

The treatment changed tempo, not the restaurant's entire music identity

The researchers used fast music at 94 or more beats per minute and slow music at 72 or fewer beats per minute. The control condition was the restaurant's regular playlist, which mixed fast and slow songs. The treatment playlists were selected from that same regular collection of Italian and Greek music, and the volume was similar across the conditions.

This design isolates tempo more cleanly than a comparison between music and silence or between unrelated genres. It also tells an owner what the result does not answer. The study did not test volume, lyrics versus instrumental music, genre, musical mode, silence, or whether music matches the concept. Changing those variables would be a different experiment.

The study took place in one busy Italian restaurant in central Tel Aviv

Data were collected over several months during evening hours at one Italian restaurant in central Tel Aviv. The restaurant had been operating for approximately 20 years, served a mostly middle-class and above clientele, and was described as very busy throughout the week. The study used randomly assigned days for the three conditions and observed patrons in their natural dining environment.

That setting gives the evidence practical value, but it also limits generalization. A fast-casual lunch restaurant, a quiet fine-dining room, a bar, and a family restaurant may respond differently. The study also could not measure subjective reactions because the researchers avoided questionnaires so that patrons would behave naturally.

Test music against table capacity, not against a promise of higher sales

An owner can use the study to design a local test. First record the baseline average table time, average bill, tip percentage, table utilization, queue length, and guest-satisfaction score by daypart. Then test a tempo change only in comparable service periods while keeping prices, menu, staffing, and promotion as stable as possible.

Useful calculations include:

The music study provides an observed duration difference and a controlled tip-amount result. Your POS, reservation, waitlist, and satisfaction data must establish whether a change in tempo actually improves the outcome you care about. If the dining room is not demand constrained, a shorter stay may not add revenue. If guests dislike the atmosphere, faster turnover may not be worth the trade.

Sources