Consulting Case Interviews for Campus Hires - Structure Over Cleverness

Key Takeaways

• Cases assess structured thinking under pressure, not business knowledge. You are not expected to know the industry.
• Memorised frameworks are visible and score badly. A structure built from the specific question scores well.
• Arithmetic is done aloud. Stating your approach before calculating protects you when you make an error.
• The fit interview is not a warm-up — for close calls it frequently decides the outcome.
• Consulting recruits earliest of any campus industry, so the timeline matters more than in most sectors.

What a case is actually testing

An interviewer gives you an ambiguous business problem: a client's profits are falling, should they enter a market, is this acquisition sensible. You have twenty to thirty minutes.

They are scoring four things:

Structure. Can you break a vague problem into parts that are separate and that together cover the question?

Prioritisation. Having built a structure, do you know which branch matters most, and can you say why?

Quantitative comfort. Can you set up and execute arithmetic without a calculator, and sanity-check the result?

Communication. Can you be followed? Do you signpost where you are? Do you say what the number means?

Notice that business knowledge is not on the list. Cases are set in industries you have never worked in, deliberately. If you find yourself thinking you cannot answer because you do not know the airline industry, you have misread the exercise.

Why memorised frameworks fail

Every case book teaches frameworks and every firm's interviewers can hear them being recited. Understanding why they fail changes how you prepare.

They are audible. An interviewer who has run four hundred cases recognises a framework announcement instantly. It signals a candidate reproducing something rather than thinking.

They do not fit. A market-entry framework applied to a profitability problem produces branches that are irrelevant to the question. The interviewer then watches you explore them, which wastes half the case.

They are exhaustive rather than prioritised. The whole point of structure is deciding what matters most. A framework that lists everything demonstrates the opposite skill.

They collapse under a curveball. When the case does not match any framework — and interviewers deliberately choose cases that do not — the candidate with nothing but frameworks has nothing.

What to do instead. Learn the underlying logic rather than the shapes. Profit is revenue minus cost, revenue is price times volume, costs split fixed and variable. That much is arithmetic and worth internalising completely. Then build the branches from the specific facts you were given.

The tell that you are doing it right: your structure contains words from the prompt. "Coffee chain", "eighty stores", "two years". A structure that could be read aloud to a different case is a framework; one that could not is analysis.

Frameworks are still useful in preparation. They teach you what tends to matter in a profitability or market-entry problem. Use them to build intuition, then leave them behind in the room.

Handling the maths

Case arithmetic is not hard maths. It is simple arithmetic done aloud, under observation, with no calculator — which is a different skill.

State the approach before you calculate.

"So I want annual revenue. That's units per month, times twelve, times average price. Units are 40,000, so 480,000 a year, times $25 — call it $12 million."

If you make an arithmetic slip, the interviewer has seen that your method was right and will usually just correct the number. If you calculate silently and produce a wrong figure, they have nothing to credit.

Round aggressively and say you are doing it. "I'll use 500,000 rather than 483,000 to keep it clean" is fine and expected.

Sanity-check out loud. "That gives a 40% margin, which seems high for retail — worth checking whether I've missed a cost line." Noticing an implausible number is a strong signal.

Practise mental arithmetic separately. Percentages, ratios, and multiplying and dividing large round numbers. Ten minutes a day for a few weeks removes most of the difficulty.

Structuring the whole case

Clarify first. Repeat the question in your own words and confirm the objective. "So we want to know whether to enter the German market, and success means profitable within three years — is that right?" This takes twenty seconds and prevents answering the wrong question.

Ask for time. Thirty to sixty seconds to structure is normal and expected. Say you are taking it.

Signpost constantly. "I've looked at revenue and it seems stable, so I'd like to move to costs." The interviewer should always know where you are.

Drive. Cases are collaborative but you are expected to lead. Waiting to be told what to examine next scores badly.

Finish with a recommendation. Not a summary — a recommendation, with the reason, the main risk, and what you would check next. Even when you are not certain, commit. "I'd recommend entering, primarily because the margin structure holds at half the volume we assumed. The main risk is the regulatory timeline, and I'd want to pressure-test that before committing capital."

Trailing off without an answer is one of the most common ways strong analysis produces a rejection.

The fit interview

Consulting fit interviews are more demanding than most students expect. They are looking for leadership, resilience, working with difficult people, and persuading without authority — and they probe hard.

Expect follow-ups: why did you do that, what did the other person say, what would you do differently, how did you know it worked. A story that holds up for ninety seconds but collapses under three follow-ups is a common failure.

The preparation is the same as any behavioral story bank, with more depth per story. Six well-examined stories beat twelve shallow ones.

Timing and process

Consulting recruits earliest of any industry on campus. Applications frequently open in July and August for the following year, with offers out by autumn. Some firms run sophomore programmes, meaning the pipeline begins in your second year.

If consulting is your target, the autumn window is not advice, it is the entire game. Missing it usually means waiting a full year.

The process typically runs: resume screen, sometimes an online test, first round of one or two cases plus fit, then a final round of two or three more with senior staff.

The resume screen is real and unforgiving — these firms receive enormous volume and cut hard on it. Your document needs to survive automated parsing and read as achievement-dense rather than duty-dense. Tailoring it to the firm matters, and so does having something concrete on it beyond coursework.

How to practise

Do live cases with a partner. Cases cannot be practised alone — the interaction is the exercise. Case clubs exist at most universities for exactly this reason.

Do fewer, deeper. Thirty cases with real feedback beats eighty rushed ones.

Get feedback on communication, not just conclusions. Ask your partner whether they could follow you, not whether you were right.

Record one. Uncomfortable, and the fastest way to notice that you mumble through arithmetic or never signpost.

A case, walked through

Structure advice is hard to apply without seeing the shape of a real answer. This is a compressed version of a standard profitability case.

The prompt: "Our client is a regional coffee chain with 80 stores. Profits have fallen 20% over two years while revenue has stayed flat. What is going on?"

Weak opening:

"I'd like to use a profitability framework. Profit equals revenue minus costs. On the revenue side we have price and volume, and on the cost side we have fixed and variable costs. I'd also want to look at the 4Ps and consider the competitive landscape."

This is a memorised structure announced rather than a structure built. It is generic, it names frameworks that do not fit, and it tells the interviewer nothing about how the candidate thinks.

Strong opening:

"Let me make sure I have the facts. Revenue flat, profit down twenty percent over two years, eighty stores, regional. Is that profit in absolute dollars, and is the store count stable over that period?"

[Interviewer: yes, absolute dollars, store count unchanged.]

"That's useful — because revenue is flat and store count is flat, this is almost certainly a cost problem rather than a demand problem, though I want to verify that rather than assume it.

I'd look at three things. First, confirm the revenue picture is genuinely flat rather than two offsetting movements — say, falling volume masked by rising prices, which would change the diagnosis completely. Second, break costs into the big buckets for a coffee chain: cost of goods, labour, occupancy. Third, check whether this is broad across stores or concentrated in some of them, because eighty stores means an average can hide a lot.

I'd start with the second one — the cost breakdown — since the fact pattern points there. Do we have cost data by category?"

Same opening thirty seconds, entirely different signal. It clarifies before structuring, it draws an inference from the given facts, the structure is built for this problem rather than pulled off a shelf, and it ends by prioritising and asking for the next piece of data.

The four moves in that answer, which generalise to every case: clarify the facts → draw one inference from them → lay out a small number of specific areas → say which you would do first and why.

Handling the arithmetic

Where competent candidates most often come apart, and it is a training problem rather than an intelligence one.

State the approach before you calculate. "So if labour is 30% of a $40 million cost base, that's $12 million, and a 10% wage increase adds $1.2 million." Saying the method first means an arithmetic slip is a slip rather than a failure — the interviewer can see you knew what you were doing.

Round aggressively and say you are doing it. "Let me call it 80 million rather than 78.4, and I'll adjust at the end if it matters." Nobody wants four significant figures; they want speed and judgement.

Sanity-check out loud. "That gives an average of $340,000 per store per year, which feels plausible for a coffee shop — call it a thousand dollars a day." This catches the order-of-magnitude errors that actually sink candidates, and it demonstrates business sense.

Ask for a moment rather than talking through the sums. "Give me twenty seconds to work through this" is completely normal. Narrating long division is not.

Practise mental arithmetic separately from cases. Ten minutes a day of percentages, fractions and multiplication of round numbers. It is a distinct skill from case structuring and it is the one that fails under pressure.

Write legibly and keep your page organised. Interviewers glance at your paper. A structured page reads as a structured mind; scattered numbers read as panic.

Structuring the whole case, not just the opening

The opening structure gets all the attention, and cases are lost in the middle.

Signpost constantly. "I've looked at costs and found labour is the driver. I'd now like to understand why labour costs rose — is it wage rates, hours, or headcount?" The interviewer should never be unsure where you are.

Drive the case. Ask for the data you need rather than waiting to be handed it. Passive candidates who answer each question and stop score badly even when every individual answer is right.

Say what a finding means before moving on. Every number you receive should be followed by one sentence of interpretation. Data without interpretation is a candidate doing arithmetic, not analysis.

Take the interviewer's hint. If they say "that's interesting, but what about the revenue side?", go there immediately. Hints are not traps and refusing them is costly.

Manage the clock. A case is twenty to thirty minutes. If you are fifteen minutes in and still exploring, you will not reach a recommendation, and a case with no recommendation fails regardless of the quality of the analysis.

Finish with a recommendation, out loud, in structure. "My recommendation is X. The reason is Y. The main risk is Z, and the thing I'd want to check next is W." Even if you are unsure. A confident, structured, imperfect recommendation beats trailing off, every time.

The first ninety seconds

The opening of a case sets the tone for everything after, and it follows a fixed sequence worth drilling until automatic.

Play back the facts. One sentence confirming what you heard. It catches misunderstandings before they cost you fifteen minutes and it buys thinking time.

Ask one or two clarifying questions. Not five — a barrage reads as stalling. Good ones establish the objective and the metric: what does the client mean by success, and over what horizon.

Ask for a moment to structure. "Could I take thirty seconds to organise my thinking?" Completely standard, and candidates who plough straight in produce worse structures.

Then lay out your approach, with the prompt's own words in it.

Three or four specific areas, not a framework name.

Say which you would start with and why. This is the part most often skipped, and prioritisation is the competency being assessed.

Ninety seconds, five moves. Practise the sequence separately from full cases until it is automatic, because it is the part of the case where nerves do the most damage.

Common Mistakes

• Deploying a memorised framework. Visible, and it produces irrelevant branches.
• Silent arithmetic. Removes the interviewer's ability to credit your method.
• Not driving the case. Waiting for direction scores as passivity.
• No recommendation at the end. Analysis without a conclusion reads as an inability to commit.
• Ignoring fit preparation. Frequently the deciding factor between similar candidates.
• Starting in the spring. Consulting closes earlier than any other campus industry.

Practising properly

More cases is not the answer. Better cases is.

Do live cases with a partner from the start. Reading case books alone builds none of the skill being tested, which is thinking aloud while someone watches. Twenty live cases beats sixty read ones.

Take turns interviewing. Giving a case teaches you what the interviewer sees — where candidates ramble, where they lose the thread, where a structure stops being useful. Most people improve more from giving ten cases than from taking ten.

Do fewer, deeper. Repeating a case you handled badly a week later, properly, teaches more than a new one. Candidates who count cases plateau.

Get feedback on communication, not just conclusions. "Did I get the right answer?" is the least useful question. Ask: was my structure clear, did I signpost, did I drive, did I interpret the numbers, did I sound confident when uncertain.

Record one and watch it. You will discover the filler words, the long silences, and the fact that you never said what the numbers meant.

Around fifteen to twenty-five live cases is the range where most candidates become competent. Beyond forty the returns are minimal and the risk of sounding rehearsed rises.

Practise the fit interview at the same time. It is not a warm-up — for close calls it frequently decides the outcome, and candidates who prepare cases exclusively arrive with nothing for "why consulting" beyond generalities. Build the story bank alongside the cases rather than after them.

The fit interview, and why it decides close calls

Nearly everyone reaching a final round can do the cases. What separates them is the other half.

"Why consulting?"

The single most important question, and the one most often answered with the reasons consulting firms put in their own marketing — variety, learning, smart colleagues. Those are true of many jobs. A strong answer connects something you actually did to what the work involves: "I spent last summer trying to work out why our volunteer retention was terrible, and the part I enjoyed was that the answer was not the one anyone expected. I want more problems where the diagnosis is the hard part."

"Why this firm?"

Requires real research and ideally a conversation. Naming an office, a practice area, or something a person told you separates you from the majority who answer interchangeably.

"Tell me about a time you led." Leading without authority is the relevant version, since that is what consultants do constantly.

"Tell me about a failure." Testing self-awareness. Disguised strengths score zero.

"Tell me about a time you influenced someone." Central to the job and frequently asked.

Interviewers compare notes across the day, so consistency matters — a story that grows between the morning and afternoon rounds gets noticed. Prepare six stories, use them across interviewers without worrying about repetition, and keep the details identical.

Timing, and what the process looks like

Consulting recruits earlier than almost any other campus industry, and the calendar is the part that catches students out more than the cases do.

Applications open in the summer before your final year, with most deadlines in July and August for autumn interviews. Students who begin thinking about consulting in September have generally missed the main round.

Networking events run from spring onwards and at several firms attendance is tracked. This is one of the few industries where showing up at events measurably affects your application.

The process: application and resume screen → online test or digital assessment → first round of two cases with fit → final round of three to five interviews, frequently including a partner. The whole sequence often completes within six weeks.

Offers arrive with deadlines that are shorter than in most industries, and they sometimes expire before other sectors have finished interviewing. Asking for an extension is possible and less often granted here than elsewhere.

Undergraduate and advanced-degree tracks are separate, with different timelines and sometimes different processes. Applying to the wrong one wastes an application.

What this means practically: if consulting is your primary target, the preparation starts in the spring of your penultimate year — cases from around March, applications ready by June. Starting in September means competing against people who have done thirty live cases with a partner they have been practising with since April. That gap is visible in the room and it is the single biggest predictor of who converts.

Because the calendars barely overlap, hedging across consulting, banking and technology generally produces three under-prepared campaigns. Pick a primary track early, prepare properly for it, and treat the others as genuine backups rather than parallel efforts.

Frequently Asked Questions

Do I need a business degree?

No. Firms hire across disciplines deliberately, and engineers, scientists and humanities students are all recruited. Business knowledge is not what is being tested.

How many practice cases should I do?

Enough that structuring stops feeling like a scramble — commonly twenty to forty with live partners. Depth of feedback matters more than the count.

What if I get stuck mid-case?

Say so and reason aloud. "I'm not sure which of these matters more — let me think about which would move the number most." Interviewers help. Silence is what hurts.

Is mental maths really that important?

Yes, and it is the most trainable part. Percentages, ratios and large-number arithmetic, practised briefly and daily, remove a whole category of avoidable failure.

Does my school matter?

It affects access to on-campus recruiting more than it affects assessment. Where firms do not visit your campus, referrals and direct applications replace that channel — cold outreach does real work here.

Should I write while I think?

Yes. Structure your notes so you can glance down and see where you are. Interviewers expect note-taking and often look at your page.