← Back to Work

To Truck or not to Truck?

Timeframe2025
ProjectMr. Bulgogi
RoleData Analyst
ToolsGoogle Sheets · NPV Modeling · Survey Research

Summary

Mr. Bulgogi is a Korean food truck that shows up on Carnegie Mellon's campus a few weeks a semester and reliably sells out. For a Decision Analysis course, I worked with a team of four to answer the question CMU Dining Services actually has to answer: should Mr. Bulgogi stay a seasonal pop-up, become a permanent restaurant, or stay a pop-up while a competing Korean restaurant opens permanently instead? I led the data side, designing the student survey, building the Even Swaps and NPV models, and pressure-testing the recommendation against the assumptions it depended on.

Mr. Bulgogi's red 'Tartan Express' food truck parked on Carnegie Mellon's campus

The Decision

The decision sits with CMU Dining Services, weighing three real objectives at once: student satisfaction, building a base of regular customers, and profit. Three alternatives were on the table:

Go permanent

Establish Mr. Bulgogi as a full-time, on-campus restaurant.

Stay the course

Keep Mr. Bulgogi exactly as it is: a seasonal food truck, nothing added.

Pop-up + competitor

Keep the truck seasonal, but let a new Korean restaurant open permanently alongside it.

Turning Opinions Into Dollars

We ran a 104-respondent student survey and went directly to the sources for operating data, emailing Mr. Bulgogi's owner for cost figures and CMU Dining Services for contract and space-availability context. Student satisfaction isn't naturally priced, so we used Even Swaps to convert it into dollar terms: we asked what students would be willing to pay for a bowl, for eating weekly, for the option to eat there again, and for a "significant," "moderate," or "slight" improvement in satisfaction, then applied those rates against each alternative's projected traffic. That let satisfaction sit inside the same financial model as rent, ingredients, and revenue, rather than being a bullet point next to it.

Each alternative's monetized cash flows were then projected over a 20-year horizon and discounted using CMU's own weighted average cost of capital (7.01%), calculated from the university's 2024 financial statements: the same discount rate a university treasurer would actually use, not a generic assumption.

What the Model Found

Going permanent wins by a wide margin. Its 20-year NPV comes in roughly ten times higher than either staying seasonal or ceding the permanent slot to a new competitor, driven less by the extra open days themselves than by unmet demand the survey surfaced directly: over half of respondents said they wanted Mr. Bulgogi but it simply wasn't open when they were looking, and a majority said a full-time location would meaningfully improve their satisfaction with campus dining.

Chart comparing 20-year NPV across the three alternatives, and key survey findings from 104 respondents

Stress-Testing the Recommendation

A model is only as trustworthy as its shakiest assumption, so we built two-way sensitivity tables (discount rate against improvement rate, and improvement rate against carryover rate) spanning a wide range around our base-case estimates. Going permanent stayed the top choice in every single combination we tested. That mattered more than the headline number: it meant the recommendation didn't depend on us guessing the exact right inputs, just the right direction.

Recommendation

We recommended that CMU Dining Services open Mr. Bulgogi as a permanent, full-time restaurant on campus. It's projected to substantially outearn either alternative over a 20-year horizon, the result holds up under a wide range of assumptions, and it directly answers demand the survey data made hard to ignore: students who already wanted this and couldn't get it.

Assumptions & Key Uncertainties

What the model assumed: that a full-time location sells more bowls per day than the truck does (based on the share of survey respondents who said they'd eat more often if Mr. Bulgogi were open full-time, applied as an improvement rate over current weekly demand); that a new competing restaurant would land at roughly 90% of Mr. Bulgogi's satisfaction with students, with some of that demand carrying over rather than being lost outright; and that total revenue scales with bowls sold times customers, with some natural variability from add-ons, appetizers, and desserts.

What we couldn't fully resolve: whether students are actually as unsatisfied with current dining as the survey suggests, versus just enthusiastic about a truck they already like; whether a permanent location would genuinely raise satisfaction or just formalize an existing favorite; how daily order volume really shifts between a temporary and permanent format once the novelty wears off; the risk that a permanent Mr. Bulgogi cannibalizes traffic from other campus dining options rather than growing the pie; and how much of the revenue CMU actually keeps, since contract terms and margin splits with dining partners weren't fully available to us.