McDonald's Drive Thru
Service Redesign
About the Project
McDonald's is going through a digital transformation. Digital kiosks are replacing cashiers. McDelivery phone number is now an app. With the app update in 2019, we see consolidation of McDonald's services and increased engagement of McDonald's and their users, especially with the new option to redeem deals by scanning barcodes from our phones. However, the drive-thru is still running on the legacy system.
(L-R) Drive Thru for deposits at Grand National Bank of St. Louis, First fast food drive-thru in 1948, first McDonald's drive-thru was created in 1975 near Fort Huachuca
History of Drive-Thrus
The first recorded use of a drive-thru was in 1930 as a drive-up window teller at Grand National Bank of St Louis, Missouri. In 1948, the first fast-food drive-thru was opened when In-N-Out started as a Drive-Thru only restaurant chain.
In 1975, McDonald's opened the first drive-thru to serve military members who were not permitted to leave their cars while wearing fatigues. Created out of a need than convenience, it consisted of a microphone and orders picked up in person at the window - the same system used currently.
UX Challenges
Understanding Happy Users
To understand the challenges faced by users of the drive-thru process, we launched our research consisting of usability interviews, contextual inquiry and service safari. Customers were often happy with the service for its convenience. However, it was only through contextual inquiry that I was able to see opportunities for improvement. Some of these observations were:
-
Summary order screens were often ignored
-
Miscommunication happened often
-
Customers had different opinions and concerns on the best payment method
These were grievances that customers forget once they received their meals.
Space constraints
We launched a service safari to understand the route. Unlike the US, space constraint is integral to the issue - both for drive-thru customers and for service. This led to:
-
long queues resulting in customers who skipped the drive-thru altogether
-
incremental change was not possible for service. As the drive-thru had to be near the window. In addition to structural challenges, the restaurant also faced manpower issues. This compounded with service standards they tried to achieve (120s from order to receiving of order,) and their preferred payment.
We had the luck of observing an off-peak order. The service standards were unfortunately not met. We also considered the challenge of testing and reiterating the prototype. How can we simulate a traffic jam? How can our improvements be made measurable?
Drive-thru walk through. Animation by Zakhran Khan
Service
Blueprint
From the interviews and user research, I created the service blueprint to see opportunities within the service and to understand how one change could affect other parts of the service. We narrowed to two areas of opportunities as highlighted in yellow.
UX Solutions
We targeted our solution for the McDonald's Drive-Thru at Springleaf Towe. It was the lowest common denominator in terms of space and parking. We bridged existing digitization efforts of McDonald's with the current process through the
-
introduction of digital order board to improve accuracy
-
an in-app ordering system, which allows users to generate a barcode for their orders
-
combination of payment and collection counter to reduce manpower and allow the possibility for vehicles with large orders to move ahead to park while waiting for their orders
Bodystorming #1
To work around the short time frame and bureaucracy, we decided to test our prototype via bodystorming sessions. During our dry run, the lack of feedback created a dissonance. How can we provide signifiers and confidence that their order has been taken by the new system? We decided to issue a stub and complete the order with an automated voice message.
Lesson
The queue at the digital ordering board occurred faster than usual and testers did not manage to give the in-app ordering system a chance. Even though it was digitized, we found that
-
physical signifiers such as "Order Counter", "Payment Counter" and physical suggestions of the payment methods were necessary for the success of the bodystorming
-
testers would abandon the app once they were at the ordering kiosk. They asked if it is possible to continue ordering at the kiosk if it was a wrong order
-
more system feedback is required from the ordering system such as "Added to Cart"
Left: 1st Thank you screen
Right: Final Thank you screen
esters were often confused by the map and overlooked their ticket stub.
.png)
Bodystorming #2
We put in place physical signifiers, increased voice feedback, number of kiosks, we had testers travel in twos to simulate driver and passenger.
We found users to be agitated when the person in front of them was slow and we realised that managing customers expectations were necessary for creating happy customers. Additionally, the extra kiosks and payment machines were often underutilized.
Final Solution
-
We proposed a single ordering kiosk and a single cashier with a parking lot for big orders that might require more time
-
We provided more signifiers so that customers are able to generate a barcode and continue their order at the kiosk instead of abandoning it
-
Customers would also be able to save their favourite orders to allow a quick ordering possibility as users have also reflected that they tend to order the same things
Next Steps
I would propose for this to be taken for a trial with the use of a tablet in the presence of a staff who will report the orders manually to the kitchen. A barcode scanner connected to the tablet can also be put in place for users to first scan in the deals they would like to redeem. This should precede the introduction of the possibility of an in-app ordering system.
Launch
Once the system is launched, staff should ideally be present to assist customers with their orders for the first few months.
Voice
We considered the use of voice recognition systems. However, being in a multilingual country and the nature of food order (read: cancellations, customisations), an ordering system would need to be created to make it more user-friendly and accessible.
Team
Gwen Sia
Nurulhuda Izyan
Rachel Lee
Zakhran Khan
Role
Researcher
Protoyper
Interaction Designer