Research summary · Field experiment · 2006
Lavender and lemon scents in a restaurant: effects on time spent and spending
In a small French pizzeria, diners stayed longer and spent more on a Saturday evening scented with lavender than on an unscented Saturday. A lemon scent made no significant difference to either.
Reviewed by Simon Faure-Field · 8 October 2026 · Based on published accounts of the paper
At a glance
Key findings
- 91 → 106minutes: average time diners stayed, without scent and with lavender.
- €17.5 → €21.1average spend per diner, without scent and with lavender.
- 88diners across three Saturday evenings.
The study
Citation and study details
Citation: Guéguen, N., & Petr, C. (2006). Odors and consumer behavior in a restaurant. International Journal of Hospitality Management, 25(2), 335–339.
Read the record: Research record (HAL / RePEc)
- Study typeField experimentPeer-reviewed
- Sample88 dinersThree Saturday evenings
- SettingA small pizzeriaFrance
- Evidence gradeA: direct field evidenceTime and spend measured
Method
What the researchers did
The restaurant ran three comparable Saturday evenings: no scent, lemon and lavender.
The researchers worked with a small independent pizzeria in France. On three Saturday evenings, the restaurant had no added scent, a lemon scent and a lavender scent. Lemon was chosen as a scent generally considered stimulating, and lavender as one considered relaxing. The authors reported no significant differences in the weather or the restaurant’s customers across the three evenings.
For each of the 88 diners, the researchers recorded how long they stayed and how much they spent.
Results
What they found
Lavender was linked to longer stays and higher spending; lemon was not.
With lavender, the average time diners stayed rose from 91.3 minutes to 105.7 minutes, an increase of around 15%.
Average spending per diner rose from €17.5 on the unscented evening to €21.1 with lavender. With lemon, spending was €18.1, which was not significantly different from the unscented evening, and time spent was not affected.
The authors read the result as consistent with lavender’s relaxing character: relaxed diners stayed longer, and spent more while they did.
Limitations
Limitations
Three evenings in one small restaurant.
Each condition ran on a single evening, so differences between the evenings themselves cannot be fully ruled out, even though the authors reported similar weather and customers. With 88 diners across three evenings, the sample is small.
We could not find a published abstract for this paper in the records we checked. This summary is based on published accounts of the study, including its inclusion as a real-world service-setting study in the Roschk et al. (2017) meta-analysis. We have not yet reviewed the full paper.
A note on the evidence
How strong is this evidence?
Suggestive: a widely cited real-world result from a small, simple design.
This is one of the few studies to measure actual time and spending in a working restaurant, which is why it is so often cited. Its design is simple and it has not been directly replicated, so its specific figures should be treated as indicative. Its direction is consistent with later evidence that scent effects tend to be stronger in service settings.
For commercial spaces
What it means for commercial spaces
For a restaurant or café, the study suggests a relaxing scent may encourage guests to linger and order more. That is useful on quieter evenings, but longer stays can work against a restaurant that needs to turn tables at peak times. The scent should suit the setting and sit well alongside food aromas, and it should be tested over more than one evening before drawing conclusions.
Scent Swirl supplies ambient scenting for hotels, restaurants and other hospitality venues. See how scenting works in hospitality →
How to cite this summary
Scent Swirl. (2026). Summary of Guéguen & Petr (2006), Odors and consumer behavior in a restaurant. Scent Research Library. https://scentswirl.com/research/gueguen-petr-2006-lavender-restaurant/ (last reviewed 8 October 2026 by Simon Faure-Field).
This summary is based on published accounts of the paper. We have not yet reviewed the full paper. If you spot an error or know of a study we should add, we would like to hear from you.
Keep reading