Cameron Erdogan


User Research + Contextual Inquiry + Service Design

A case study on improving the campus lunch experience.


User Research + Contextual Inquiry + Service Design

A case study on improving the campus lunch experience.


This project involved using research-driven design to understand and improve lunch at CMU. We chose to focus on a unique phenomenon at CMU: lunch box delivery from Chinese restaurants near campus. After conducting a contextual inquiry to better understand the process and stakeholder values, we diagrammed our insights and generated designs based on observed pain points. MealBox is the resultant design from this process. 



Usually the line is much longer than this (I swear)

Usually the line is much longer than this (I swear)

Our design team was given the broad task of improving the lunch experience for students on CMU’s campus. We were given free rein to decide how exactly we would do so. After an initial brainstorming session in which we mapped out different parts of lunch on campus, a particular phenomenon struck our interest: Chinese-speaking students use WeChat (a mostly Chinese messaging app) to order take out lunch from nearby Chinese restaurants. At lunch time, everyone who ordered lunch meets with a delivery driver parked outside of a campus building to pick up their lunch and pay

This process is notable because it succeeds in spite of the limitations and rudimentary technology used throughout the process. For instance, the individual WeChat groups are each managed by a different restaurant owner, who tally orders by hand. Additionally, there are only 1 or 2 different lunch options from a restaurant on a given day.

This is a unique phenomenon that I hadn’t experienced before I came to CMU. It seems to reveal values and motivations in students that are otherwise underserved. Thus, it is a fertile subject to explore. 


Contextual Inquiry Process

After deciding to focus on Chinese restaurant lunch boxes as our area of focus, our first step was to perform contextual interviews with stakeholders. Contextual interviews are useful when you want to learn about a process that you know little about. In a contextual interview, the interviewer observes the interviewee while they are performing the task in question. When the interviewee does something the interviewer doesn’t understand, the interviewer asks a question to clarify and learn. A common analogy is that the interviewee is the “master” and the interviewer is the “apprentice”. 

In this case, it contextual inquiry means interviewing different lunch-eater archetypes and restaurant owners. Our five person team split up and performed a contextual interview with four lunch-eaters and one restaurant owner/chat group operator. Throughout the interview, each of us took notes that we later compiled and shared with the other group members.


Model Construction

As a group, we went through notes from each interview one by one. We generated a sequence-flow model diagram and a cultural model diagram to map out our understanding of how each interviewee went through ordering/picking up/preparing lunch. Later, the five sequence-flow and cultural diagrams were each combined into a consolidated model. These models are shown below.

consolidated sequence-flow model

consolidated cultural model

We also used our collection of interview notes to construct an affinity diagram. The goal of the affinity diagram is to organize the notes into a hierarchy of themes. From these themes, patterns and design ideas emerge. A few selected chunks of our affinity diagram are shown below (the full diagram is quite huge).



A few interesting things came up during this process. The most significant thing to note is that nearly all of the people who use this are, unsurprisingly, Chinese. These students felt there weren’t many good Chinese food options on or near campus, so picking up a take-out box from a delivery car was acceptable, since it was the best and only way to get the food they wanted. In fact, as of now, only Chinese restaurants appear to take this approach to selling lunch.

Additionally, some of things we might’ve thought of as “hurdles” weren’t so bad in many cases. The delivery restaurants have a limited menu (often only one type of item a day), but students have plenty of time before actually lining up to decide if they’re interested. The freshness of the food also wasn’t an issue for students (the items are stored in coolers in the back of an SUV). 

An overwhelming feeling among students was that they were busy and didn’t have much time to eat lunch. Waiting in line for 10 minutes to pick up a modestly priced,  ready-to-eat takeout lunch you know you want to eat was very appealing. 


Design Generation

The affinity diagram, consolidated cultural diagram, and consolidated sequence-flow diagram were used for a design exercise called “walking the wall”. In this exercise, the diagrams are annotated with post-it notes that indicate design ideas, questions, and breakdowns. These notes serve as the basis for informing and generating designs. We performed various group brainstorming techniques to think of new (and weird) ideas in the lunch box space. Some activities included bodystorming, storyboarding, and “yes, and” exercises.

We then selected a few of our most promising design ideas and made storyboards for them. There storyboards were shown to potential users in “speed dating” sessions for need-validation purposes. The idea here is that by showing your target user a vision of a preferred future, you can validate the user need that your research supposedly uncovered. After speed dating, we decided on a design and refined it once more.


Final Design

Of the various needs we identified through user research, the most compelling was this: students will deal with pre-made lunch if it’s convenient, quick, and something they can’t easily get on campus. Our proposed solution designs around this need. MealBox addresses the breakdowns in the current lunch box delivery process, effectively lowering the barrier of participation for new users.


Among the many features, the most notable is the MealBox locker. Delivery people deposit lunches into a locker at lunch time. The locker notifies the student that their order is ready to be picked up, and keeps the meal hot and fresh in the meantime. In our design, there will be multiple MealBox locker clusters placed at various convenient points on campus. The ordering and pickup happens via an app. 

This is of course, just a design at this point. It's worth noting that these types of hot meal storage units are not some far-out future technology. They exist, mostly in European convenience stores to keep hot food fresh. The most important innovation conceived by me and my team is the connection of these yet unconnected technologies.