The 2007 Drive-thru Performance Study was produced and conducted by Insula Research. This research was supported by sponsors QSR magazine, Phase Research (founding sponsor), First Data Corp, VeriFone, and Archon Development (Prophet).
The primary objective for this study was to determine the leading restaurant chains in several categories of drive-thru operation. Categories measured in the study include but are not limited to: wait time, service time, order accuracy, suggestive-sell incidence, exterior appearance assessment, speaker clarity, menuboard appearance, and availability of electronic payment options and order-confirmation boards.
Data for the study was collected during the months of April, May, June, and July 2007. Restaurants were visited between the lunch hours of 11:00 a.m. and 2:30 p.m. and between the dinner hours of 4:00 p.m. and 7:00 p.m. A restaurant could only be visited twice—once during the lunch period and again during the dinner period. A total of 10,355 service times were collected, while a total of 5,177 orders were reviewed for accuracy. The study included assessments from restaurants in all 50 states and included 25 quick-serve restaurant chains.
Speed was measured in two different time segments: wait time, or the time from when the test vehicle entered the line to when the vehicle stopped at the order station, and service time, or the time from when the vehicle stopped at the order station to when the entire order was received, including change.
Accuracy percentages were based on the percentage of orders received by researchers exactly as placed. During each visit, researchers ordered a main item, a side item, and a drink. One special request was made per order. After receiving the order all food and drink items were checked for complete accuracy. Any food or drink item not received exactly as ordered was listed as inaccurate. Similarly, orders were considered inaccurate if a requested condiment was not received, if the order failed to contain at least one napkin or one straw, or if the cashier gave incorrect change.
Sample sizes were selected based on the total number of drive-thru units each chain operates. Analysis was performed to set sample sizes in such a manner as to result in approximately the same margin of error for each chain included in the study. For this study the maximum error (i.e. worst case scenario) at a 95 percent confidence level was targeted at approximately 5.7 percent in a two-way split of the data.