This article originally appeared on my blog in August of 2014. We regularly get inquiries on this topic so, I thought it would be appropriate to share our findings again. If you are a regular reader of my blog, then you know I am passionate about the need to collect local trip generation data and even more passionate about sharing it with others. I am including links in this article for you to access the data we have collected. If you share my passion for sharing data with other transportation professionals, please contact my colleague Nate Hood. We are always looking for data collected on interesting land uses, and would love to share your findings on our blog.
We’ve done several traffic impact studies for developments with restaurants on the ground floor that had proposed valet zones. A nagging question has been, “how much storage is needed in the valet zone to ensure vehicles won’t block the through traffic on the road?” We couldn’t find any data out there on this topic, so we decided to collect some.
We collected data at five different popular restaurants in the Minneapolis, Minnesota area that have busy valet services. Based on our study, we found a reasonable rule of thumb is to provide space in the valet zone to park one vehicle per every 70 seats in the restaurant.
Here’s how we came up with that rule of thumb –
Data Collection Procedure
Video was collected at each location with the COUNTcam system. The maximum queue each minute was observed between 6:00 p.m. and midnight at each location. The video was watched at approximately 10x speed, with frequent pauses to write the maximum queue number per minute in a matrix created and printed for this purpose. Installing the COUNTcams on Thursday and retrieving them on Monday took approximately five hours of field time to gather video at the five locations. To process the raw 60 hours of video footage took approximately 10 hours of staff time in the office.
The raw queuing data can be downloaded here.
Restaurant Details
Each restaurant and valet service provider was contacted to collect data on the restaurant and valet operations. Hennepin County property records were also reviewed to get the size of the restaurants. An important detail we learned is that the valet service providers change the valet staffing levels nightly based on the number of reservations the restaurant has.
The five locations where data was collected are listed below with descriptive information about each:
Location #1
The first location is a steak restaurant in the Uptown area of Minneapolis. Data was collected here on Friday, July 18 and Saturday, July 19, 2014. This restaurant has 70 tables (which could be modified, so the number of seats is somewhat variable), is approximately 12,600 square feet in size and it costs $7 to use the valet service. Nearby parking for patrons is extremely limited. The valet zone can hold approximately four cars.
Location #2
The second location is a steak restaurant in the downtown area of Minneapolis. Data was collected here on Friday, July 18 and Saturday, July 19, 2014. This restaurant is approximately 5,500 square feet in size. There are public parking ramps nearby with spaces typically available on weekend evenings. The valet zone can hold approximately five cars.
Location #3
The third location is an Asian fusion restaurant in the Uptown area of Minneapolis. Data was collected here on Friday, July 18 and Saturday, July 19, 2014. This restaurant has 500 seats and is approximately 5,500 square feet in size. Nearby parking for patrons is extremely limited. The valet zone can hold approximately five cars.
Location #4
The fourth location is an Italian restaurant in St. Louis Park, MN. Data was collected here on Friday, July 18 and Saturday, July 19, 2014. This restaurant has 100 tables and it costs $6 to use the valet service. Ample free public parking is available in parking ramps within 1,500 feet of the restaurant within the suburban commercial center. The valet zone can hold approximately eight cars.
Location #5
The fifth location is a Brazilian steakhouse in the downtown area of Minneapolis. Data was collected here on Friday, May 16 and Saturday, May 17, 2014. This restaurant is 10,870 square feet. There are public parking ramps nearby with spaces typically available on weekend evenings. The valet zone can hold approximately four cars.
Data Analysis
The average, 85th percentile, 95th percentile and maximum vehicle demand at each location was calculated from the minute by minute maximum queue data. The results can be seen in Table 1.
Table 1 – Valet Parking Demand (in vehicles)
| Location | Day of Week | Average | 85th Percentile | 95th Percentile | Maximum | Valet Zone Capacity |
| Location #1 | Friday | 1.2 | 3 | 4 | 5 | 4 |
| Saturday | 1.3 | 3 | 4 | 5 | ||
| Location #2 | Friday | 1.8 | 3 | 4 | 6 | 5 |
| Saturday | 2.3 | 4 | 5 | 7 | ||
| Location #3 | Friday | 4.1 | 6 | 7 | 9 | 5 |
| Saturday | 3.0 | 5 | 6 | 7 | ||
| Location #4 | Friday | 2.7 | 6 | 7 | 8 | 8 |
| Saturday | 3.0 | 4 | 5 | 7 | ||
| Location #5 | Friday | 2.3 | 4 | 5 | 7 | 4 |
| Saturday | 1.7 | 3 | 5 | 6 |
Table 1 shows that, depending on the location, Friday or Saturday may have higher queue lengths. For each location, the average queue lengths can fit within the valet zone, but the queues do occasionally extend beyond the valet zone potentially blocking traffic on the roadway.
Table 2 compares the maximum observed queue length for each location, regardless of day, against the number of seats in the restaurant and the square footage of the restaurant. Likewise, Tables 3 and 4 show the 85th percentile and 95th percentile queues, respectively. In some cases, only the number of tables was found for the restaurant. To determine the number of seats, the number of tables was multiplied by four.
Table 2 – Maximum Queue Statistics
| Location | Maximum Queue | Number of Seats | Square Footage | Max Queue per 100 Seats | Max Queue per 1,000 Square Feet |
| Location # 1 | 5 | 280 | 12,600 | 1.78 | 0.39 |
| Location # 2 | 7 | — | 5,500 | — | 1.27 |
| Location # 3 | 9 | 500 | 5,500 | 1.80 | 1.64 |
| Location # 4 | 8 | 400 | — | 2.00 | — |
| Location # 5 | 7 | — | 10,870 | — | 0.64 |
Table 3 – 85th Percentile Queue Statistics
| Location | 85th Percentile Queue | Number of Seats | Square Footage | Max Queue per 100 Seats | Max Queue per 1,000 Square Feet |
| Location # 1 | 3 | 280 | 12,600 | 1.07 | 0.24 |
| Location # 2 | 4 | — | 5,500 | — | 0.73 |
| Location # 3 | 6 | 500 | 5,500 | 1.20 | 1.09 |
| Location # 4 | 6 | 400 | — | 1.5 | — |
| Location # 5 | 4 | — | 10,870 | — | 0.37 |
Table 4 – 95th Percentile Queue Statistics
| Location | 95th Percentile Queue | Number of Seats | Square Footage | Max Queue per 100 Seats | Max Queue per 1,000 Square Feet |
| Location # 1 | 4 | 280 | 12,600 | 1.43 | 0.32 |
| Location # 2 | 5 | — | 5,500 | — | 0.91 |
| Location # 3 | 7 | 500 | 5,500 | 1.40 | 1.27 |
| Location # 4 | 7 | 400 | — | 1.75 | — |
| Location # 5 | 5 | — | 10,870 | — | 0.46 |
Tables 2 through 4 shows there is a stronger correlation between the maximum queue lengths and the number of seats than the square footage of the restaurant. This makes sense because the number of seats and tables is ultimately what determine how busy a restaurant can be.
Conclusions
Based on these results, for restaurants in the Minneapolis, Minnesota area on a Friday or Saturday evening, a valet service averages between one and four vehicles in the valet zone at almost all time. The maximum queue was approximately one vehicle per 50 seats in the restaurant. Based on the queue 85th and 95th percentile data, it is reasonable that valet parking zones should provide parking space at a rate of one space per 70 seats in the restaurant.
These numbers are from a small sample size and should not be used as design guides, but rather to give a general idea of what can be expected at a valet zone. Other factors that affect queuing at a valet include the number of people on staff at the valet, the cost to use the valet service and the availability of nearby parking.
You can review the full data and more at www.TripGeneration.org. That’s our free site to share trip generation data with everyone, with currently over 5,000 hours of professionally collected traffic data for popular land uses.
Want to partner with us to collect your own local trip generation data? We have a limited pool of COUNTcam video collection products that we’re lending for free specifically for this purpose. Contact Sales at CountingCars.com or leave your contact information in the comments section if you’re interested.
Interesting work, thanks for sharing.
Did the valet service say what their SOP was for setting staffing levels per reservation? Through experience they must have a scale for each restaurant based on distance to the storage area, etc.
Scot – the valet operators weren’t very informative. They seemed to be concerned we were competitors or something trying to get proprietary information out of them. The valet companies must have ways to work out their staffing levels based on the number of reservations and the round trip time it takes their runners to do the parking. I think our rule of thumb backs into what they try to achieve. Mike
Thanks for sharing the info Mike. Very interesting and not something (the study) I’ve come across yet.
Hi Mike,
Good work and thanks for sharing the data. Did you look at the number of valets/drivers (I might have missed it). I think if there is inadequate parking lane, adding more valets would speed up the process and reduce queuing.
Pang
We did try to get data on the number of drivers, but weren’t successful. I wrote a previous comment on it.
Do the valets use all the spaces for operations or do they use some for VIP parking? We have observed a lot of valets that operate smoothly when all spaces are used for operations but start obstructing the street as soon as a high end vehicle arrives and is parked in a valet space, reducing the space for operations.
Were the valets informed of your study in advance?
We have found they do not like to be watched or answer questions also.
Kevin – We included “VIP” cars that were parked in the zone. They would stay 10-15 minutes, but they didn’t linger much beyond that. We tried to work with the valet companies, but they really didn’t want to talk with us. We videoed the valet area with our COUNTcams, so the guys working didn’t even know they were being recorded. The COUNTcams are discrete and blend into the background when they’re across the street.
Mike
Thanks for sharing this. It’s interesting to see how a setup for valet parking can be affected by the amount of seating offered at an establishment, but it makes sense. I wonder how a third-party valet service might arrange themselves for operation, especially if they’re dealing with patrons from multiple establishments.