A Field Guide to Case Study Interviews (the Ones With a Printed Packet)

Someone hands you a stack of pages, a pen, and a clock. Nobody expects you to read all of it. They want to see what you do when you can't.

A case study interview usually starts the same way. You get a packet: a few pages on a company, some historical financials, maybe a page of projections or a summary of the proposed deal, and a question at the end. Should the fund buy this business? Is this price reasonable? Which of these two targets would you pursue? Then you get some amount of time alone with it, often somewhere between thirty and ninety minutes, before you walk an interviewer through your answer.

It shows up most in private equity recruiting, and some banks and growth funds use a version of it too. Sometimes it's a take-home with a few days and a laptop. The version this post is about is the in-room one, with paper and a calculator, because that's the one people tend to walk into underprepared.

What's usually in the packet

Packets vary a lot by firm, but most are built from the same parts:

• A business overview. What the company sells, who buys it, how it makes money, and sometimes a page on the industry and competitors.
• Three to five years of historical financials. Revenue, gross profit, EBITDA, and often capex and working capital, sometimes broken out by segment.
• Management projections. Usually optimistic, sometimes deliberately so, to see whether you notice.
• Deal terms or a starting point for them. An asking price or multiple, available leverage, and a holding period.
• The question. This is the only page that tells you what you're actually being asked to produce.

Some packets are padded on purpose. A page of industry trivia or a long product list is there partly to see whether you can tell what matters from what doesn't.

The first ten minutes

Read the question first. Before the overview, before the numbers. If the question is “would you invest at this price,” every page you read after that is evidence for or against one specific decision, and you'll read it faster and with a filter. If you read front to back, you spend the first twenty minutes absorbing background with no idea which parts you'll need.

Then skim the whole packet once without stopping to calculate anything. Flip through every page, note what's there, and circle anything that looks like it will matter: a margin that jumps, a customer that makes up a big share of revenue, a projection that bends sharply upward right after the deal closes.

Then decide how you'll split the remaining time. A rough split that works for a lot of people is about a quarter of the time reading, half on the analysis and math, and the rest on writing up the answer you'll present. If you get to the end with beautiful math and no recommendation written down, the math won't save you.

Reading the financials

You don't have time to analyze everything, so go for the handful of things that actually drive whether this is a good investment.

• Revenue growth, and where it comes from. Price, volume, new customers, an acquisition? Growth from one big contract is worth less than growth spread across many customers.
• Margins over time. Is EBITDA margin stable, expanding, or shrinking? A shrinking margin with rising revenue usually means the company is buying growth.
• Capex as a share of revenue or EBITDA. Two companies with the same EBITDA can generate very different amounts of cash if one has to keep reinvesting heavily to stand still.
• Working capital. If receivables or inventory grow faster than revenue, cash is getting tied up, and the cash flow will look worse than the EBITDA suggests.
• How projections compare to history. If the company grew 4% a year for five years and the plan shows 15% starting next year, ask yourself what changed. Often nothing did.

Write short notes in the margin as you go. “Margin up 300bps in two years, why?” is more useful later than a perfectly calculated ratio you can't remember the point of.

The back-of-the-envelope math

In a PE-style case, the math almost always comes down to a rough returns estimate. You aren't building a model. You're checking whether the deal can plausibly hit the kind of return a fund would want, and what has to go right for it to get there.

A worked example

Say the packet describes a business with $50 million of EBITDA, available at 10x, so a $500 million purchase price. Lenders will provide 5x EBITDA of debt, which is $250 million, so the fund puts in the other $250 million as equity. Ignore fees to keep it simple.

Now the five-year view. Suppose EBITDA grows to $70 million and you assume the business sells at the same 10x it was bought at. That's a $700 million exit value. Over the five years the business generates enough free cash flow to pay down $100 million of debt, so $150 million is left at exit.

• Exit enterprise value: $70m × 10 = $700m
• Less remaining debt: $700m minus $150m = $550m of equity
• Money multiple: $550m ÷ $250m = 2.2x
• IRR: a 2.2x over five years works out to roughly 17%

You won't calculate the IRR exactly on paper, and nobody expects you to. A few reference points get you close enough. Over five years, 2.0x is about 15%, 2.5x is about 20%, and 3.0x is about 25%. Over three years, 2.0x is about 26%. Knowing those cold lets you say a number confidently and move on.

The more useful part is splitting the gain into where it came from. The equity grew by $300 million. EBITDA growth of $20 million at a 10x multiple accounts for $200 million of that, and debt paydown accounts for the other $100 million. Multiple expansion contributed nothing, because you assumed the exit multiple matched the entry. That breakdown tells the interviewer you understand what the return depends on: in this case, mostly on the business actually growing EBITDA by 40%.

If you want more on getting this math right under pressure, the common LBO interview mistakes post walks through the errors people make most, and IRR, MOIC, and payback period covers how the two return metrics connect.

Run one downside

Once you have a base case, change one thing and rerun it. What if EBITDA stays flat at $50 million? Exit value is $500 million, and with somewhat less debt paid down (lower EBITDA usually means less free cash flow) you might have $175 million of debt left. That's $325 million of equity, about 1.3x, which over five years is an IRR of roughly 5%. One line like that shows the interviewer you thought about what happens if the plan doesn't work, and it usually takes two minutes.

Presenting the answer

Lead with the recommendation. “I'd invest, at this price, but only with comfort on customer concentration” is a good first sentence. Walking through the packet in order and arriving at a conclusion at minute nine is not.

A shape that works for most cases:

1. The recommendation in one or two sentences, including any condition attached to it.
2. Two to four reasons, each tied to something specific in the packet.
3. The returns math, briefly: entry, exit, the multiple of money, roughly what IRR that means, and where the return comes from.
4. The main risks, and which one worries you most.
5. What you'd want to diligence before committing. This is the part people skip, and it's the part that sounds most like someone who has actually done the job.

Keep it to five minutes or so unless they tell you otherwise. They'll interrupt with questions anyway, and a tight first pass leaves room for the conversation, which is where a lot of the scoring happens.

What the interviewer is actually scoring

The case is a stand-in for the actual job, which is reading a lot of information quickly and forming a view you can defend. So the scoring tends to look like this:

• Whether you took a position. A hedged non-answer is the most common way to do badly on a case where the math was fine.
• Whether you found what mattered. Spotting the customer concentration or the hockey-stick projection counts for more than calculating every margin.
• Whether your numbers are in the right range. Small arithmetic slips get forgiven. A 40% IRR on a deal with no growth and modest leverage does not, because it means you didn't sanity check.
• How you handle pushback. When they challenge your recommendation, they're often testing whether you can hold a reasoned view or update it for a good reason, not whether you'll fold.
• Whether you'd be easy to work with. Clear, calm, organized. They're picturing you presenting to a partner.

You can recommend passing on the deal and do very well. Plenty of good answers are “no.” What matters is that the no comes with reasons and a sense of what price or terms would change your mind.

Where people lose it

• Reading every page carefully before looking at the question.
• Taking management's projections at face value without saying so out loud.
• Assuming the exit multiple will be higher than the entry multiple and building the return on it.
• Forgetting to subtract the remaining debt when going from exit value to equity value.
• Running out of time with no recommendation on paper.
• Presenting in the order of the packet instead of leading with the answer.
• Treating a challenge from the interviewer as a sign you were wrong and abandoning the whole view.

Common questions