A few weeks ago, I posted the following real-world situation as a case competition:
You’ve been hired by an industry association representing the 3,000 colleges located in the United States. Due to COVID-19, nearly all universities in the United States will be conducting classes via video conference at the start of the new academic school year, which begins in August 2020.
A few months ago, the association sent out a survey to the college students attending their member universities. The survey asked a simple question:
Given the remote learning situation, do you plan to re-enroll and attend school in the next academic year?
25% responded “No.”
What should the association’s member universities do regarding this finding? How would you structure this problem? What are the possible implications of this finding?
First off, I want to thank everyone who participated for taking the time to post their answers to this question. The fastest way to learn is to take risks and try… and learn from the outcome and feedback.
I was hoping to select a winner who really cracked the case in terms of structuring it well. Unfortunately, none of the submitted answers met that standard. If this were a real case interview, I would not have passed any of the candidates to the next round.
So, that’s the bad news.
To be fair, this is a really hard question. It was so hard that when I first read this in the news, I had no idea how to solve it either. So, I decided to pass it along to all of you to attempt too.
(Yes, I was being lazy… I know. I feel a little guilty, but this problem caused my brain to hurt and I didn’t feel like trying to figure it out. I’m human. ?)
In addition, the entire higher education industry is struggling with this issue right now. They have hundreds of billions of dollars in revenue at stake, and they haven’t solved it either.
Also, this case differs from a “real” case in that you could ask the interviewer questions and they would have additional data to share. You’d very likely adapt your thinking as you discovered this new information.
So, at the start, you don’t get the benefit of uncovering this data. As a result, your case structuring needs to outline where you would start, why you would start there, and what other avenues of analysis you plan on following.
In other words, rather than provide an “answer” in the first few minutes of a case, you need to provide your plan to get an answer.
Case Structuring = Case Solving Plan
Finally, this case is difficult because it does not neatly fit into a “standard” case interview framework. Yes, you could use a profitability analysis framework, but it’s not clear that any financial data has been provided. Some participants made assumptions (a lot of them) around revenues and costs, but that’s a lot of guesstimating upon guesstimating.
You could kind of sort of use a business situation framework, but it’s limited. COVID-19 isn’t a new competitor. It’s not a new channel of distribution. COVID-19 is a disease. And guess what? Standard case frameworks don’t have a place for pandemics.
So, this is a tough problem.
Here’s the good news.
This is a teachable moment and learning opportunity for everyone involved. For those who participated, you can compare my feedback to your answer and learn from it.
For those who are observing, you can see two dozen case answers, compare them to what I’ll share below, and note the differences between the two.
Before I give my answer, let me provide some general feedback.
- Many participants jumped at a particular cause and solution without structuring the problem.Problem structuring refers to organizing the problem-solving approach into smaller sub-problems that aren’t quite so overwhelming.Examples:Some assumed it was poor students that declined to re-enroll for school and proposed tuition discounts or financial aid solutions.Others assumed it was international students that perhaps had visa problems and proposed government policy solutions.Many assumed it was due to safety and infection risks, and they proposed hygiene solutions.The reality is that these presumed causes are not known at the start of the case. It makes no sense to propose a solution to a problem that you don’t even know you have. It’s completely premature.
- Most answers focused on a “solution” as opposed to a process or plan to find the right solution.Remember, if you follow a poor process and get the “right” answer, an MBB firm can’t hire you because they can’t rely on you being lucky every time. If you follow a very sound, methodical process and perhaps were on your way toward getting the right answer, but you ran out of time, they could pass you to the next round. The thinking would be that you’re so systematic you would get the right answer eventually because of your process.
If you look at the various answers, you’ll see a lot of guesses, a lot of hypotheses, and only a few answers that had some limited plans on how to test the hypotheses.
Okay, with the benefit of several weeks to think about it and 20+ responses from other people to review, here’s how I would answer this case.
Survey results suggest that there may be a potential 25% decline in student enrollment for U.S. universities. This may be a major threat to the financial viability of the industry or certain players in the industry.
In my mind, there are two major issues:
1. Customer Demand: (In particular, a major shift in customer demand)
- What are the underlying issues that are reducing the intent to re-enroll?
- To what extent, if any, can these underlying issues be addressed to reverse or mitigate this decline?
2. Industry Supply:
Financial Viability of Existing Universities
- What are the financial ramifications of the probable range of enrollment?
- Can shortfalls be addressed, or will universities be financially unviable and need to cease operations?
- If students don’t go back to traditional universities, what are they doing instead? Should universities be offering these alternatives?
On the Customer Demand side, I’d start by segmenting the survey data. [Remember my case interview framework handout… when in doubt, segment!]
We know that 25% of respondents aren’t inclined to enroll. I’d like to know three things:
- What is the demographic profile of the students not enrolling? (Perhaps it’s only lower-income students, or international students, or students in their first year, but not those in their final year.)
- Is this trend impacting all universities equally, or is it disproportionally impacting one category of universities more than others? (e.g., Are Harvard students dropping out as much as students of other universities? Is it in-person universities only, and not online or hybrid-option universities?)
- Why are these 25% not intending to enroll? I would want to know their key concerns to see if they could be addressed in some way.
On the Industry Supply side, I’d want to know the financial structure of the various categories of universities represented by the association, and if it can absorb an enrollment decline of up to 25%. In particular, I’d want to know the magnitude of fixed versus variable costs. If the costs are mostly fixed, this will be a major bankruptcy threat. If these costs are primarily variable, this is a financial headache but not a “life-threatening” financial problem.
In addition, I want to know what substitutes or alternatives students see to attending a traditional university. Are they just taking a year off? Are they getting a job? Are they taking classes from an online university? From a low-cost community college near their hometown?
This is how I would structure/open a case… notice how I give no answers. I mainly organize this amorphous problem into two subcategories — in this case, demand vs. supply. Demand/supply seems to be a reasonable way to organize this problem.
I could have also adapted the business situation framework:
I would rename “Customers” as “Students” and could have collapsed “Competitors” and “Company” into a single category called “Universities.”
In the way I phrased this opening, I happen to state my structure explicitly (demand vs. supply), my key questions are in bullet points underneath, and I put my hypotheses in parentheses.
As a matter of preference, you could change the order around and, say, start with a hypothesis.
My hypothesis is that the 25% decline in intent to enroll comes from students from low-income families and international students. This decline primarily impacts third-tier in-person universities. I further hypothesize that universities have a very high fixed cost and that a decline of this magnitude would make them insolvent.
To disprove my hypothesis, I would need to look at two areas:
- Changes in Student Demand
- Industry Supply
I could then use the same “structure” as in my original phrasing.
As an interviewer, I wouldn’t particularly care which way you approached it. The key things I’m looking for are:
- that there is a structure.
- that the structure makes sense given the problem (e.g., you aren’t forcing a framework that doesn’t actually apply).
- that the types of questions you have (e.g., analyses you would run as a first-year consultant) are well reasoned (e.g., not arbitrary and not a list of every question in existence) and related to the case at hand.Note: This is hard to do without a hypothesis. Whether you’re silently hypothesis-driven and ask all the right questions and no useless questions, or you’re explicitly and verbally hypothesis-driven, I personally don’t care (some interviewers might and may penalize you for being overly implicit).In practice, if I suspect you’re being silently hypothesis-driven, I would simply ask you why you want to do that particular analysis.Example: Why do you want to segment the survey data? What are you hoping to learn? What question are you trying to answer? (As the interviewer, I would ask these questions to see if you have a hypothesis in mind.)
Thanks again to all who participated. I hope you find my explanations to be useful. I would encourage you to read/re-read them a few times, as there are some useful tips to help you in your case preparation.