Consultant interview questions and case interviews, with worked answers

How to use these questions

A case interview is a conversation, not an exam with one right number. The interviewer plays the client and gives facts when you ask for them. Practise each case aloud with a friend who interrupts you, and time yourself: most cases run twenty to thirty minutes. Then prepare two or three stories from your own work for the fit questions.

What a consultant is expected to know

• Structure: breaking a problem into parts that do not overlap and leave nothing out, then testing the most likely answer first.
• Numbers: quick mental arithmetic, rough estimates and simple business cases in ExcelAsked in 6% of consultant ads (as of 5 Nov 2026).Model a business case, size a market and clean client data quickly, with checks that catch errors..
• Storylines: slides in PowerPoint whose headlines tell the story, and presentingAsked in 1% of consultant ads (as of 5 Nov 2026).Build slides whose headlines tell the story on their own, and present them to a client's leaders. them to a client's leaders.
• Delivery: project managementAsked in 4% of consultant ads (as of 5 Nov 2026).Run a workstream: a plan by week, owners for each task, risks raised early and a deadline that holds., workshops and keeping the people affected on sideAsked in 2% of consultant ads (as of 5 Nov 2026).Keep clients, sponsors and the people affected on side, including those who did not ask for you..
• Technology consulting: how an ERP such as SAPAsked in 8% of consultant ads (as of 5 Nov 2026).The ERP many clients run; technology consultants configure it, move data into it and train its users. is put in place, how systems move to a cloud such as AWSAsked in 2% of consultant ads (as of 5 Nov 2026).One of the large clouds; cloud consultants plan which systems move there and in what order. or AzureAsked in 5% of consultant ads (as of 5 Nov 2026).Microsoft's cloud, common in companies that already run Microsoft software., and agileAsked in 3% of consultant ads (as of 5 Nov 2026).Deliver in short cycles with the client's team, showing working results early instead of one big launch. delivery.

The case interview: sizing, profits and market entry

Market sizing: how big is the yearly market for e-bikes in Germany?

What they check: a market sizing case tests whether you can reach a sensible number from common sense, step by step, and check it.

Short answer: "I would size it from the buyers: the number of adults, the share who ride a bike regularly, the share of those who own an e-bike, and how often an e-bike is replaced. That gives the bikes sold a year; times an average price gives the market in euros. Then I check the result against something I know, such as the number of bike shops."

The full answer, in four steps:

1. Choose the route. From the demand side (people who buy) or the supply side (shops that sell). Say which and why; demand is easier to reason about here.
2. Build the chain. Say each assumption as an assumption: about 70 million adults; perhaps half ride a bike; perhaps one in five of them owns an e-bike, so about 7 million e-bikes; replaced every seven years or so, that is about 1 million sold a year.
3. Put a price on it. At about 2,500 euros on average, the market is about 2.5 billion euros a year. Round the numbers as you go; the interviewer cares about the logic.
4. Check it. If Germany has about 6,000 bike shops, 1 million e-bikes means about 170 a shop a year, three a week: plausible. Say what would change the answer most, here the ownership share.

A weak answer: a number with no steps, or a long list of steps with no number at the end. Follow-up: "The client sells only to businesses. How does that change it?" Size the fleets instead: delivery firms, company bike schemes and rentals, each with its own count and replacement cycle.

Profitability: a client's profits have fallen for two years while sales are stable. Why?

What they check: a profitability case tests whether you split profit into its parts and follow the numbers to the cause, instead of guessing.

Short answer: "Profit is revenue minus costs. With sales stable, I would first check whether revenue really is stable once split by product, customer and price, because a shift in the mix can hide a fall. Then I would split costs into fixed and variable, see which grew, and compare with competitors. My first hypothesis would be rising variable costs, such as materials or energy, that were not passed on in prices."

The full answer: Revenue is price times volume, for each product and customer group: a stable total can hide cheaper products selling more and expensive ones less. Costs split into variable costs per unit (materials, freight, commissions) and fixed costs (rent, staff, IT). Ask for the numbers by year and find the line that moved most. Then ask why: a supplier's price, a new warehouse, discounts given by sales. Check whether competitors show the same fall; if they do, the cause is the market, and the answer is about adapting; if not, it is the client. End with two or three actions and what each is worth.

A weak answer: listing every possible cause without asking for data. Follow-up: "Costs per unit are flat, but discounts have doubled." Find who grants them and why, check whether the discounted customers are profitable at all, and propose discount rules with an owner.

Market entry: a garden tool maker from Germany wants to enter the US. Should it?

What they check: a market entry case tests whether you weigh the market, the competition, the client's strengths and the economics, and then decide.

Short answer: "I would answer four questions: is the US market attractive, meaning large, growing and profitable enough; who competes there and on what; can our client win, with its products, brand and costs; and how would it enter, alone, with a partner or by buying a company. Then I would compare the expected profit with the investment and the risk, and recommend yes, no, or yes in a smaller first step."

The full answer: Size the market and its trend first. Look at the competitors: their share, their prices, the retailers they sell through. Then the client: what makes its tools better, and whether that matters to American buyers; whether the product needs changes for the US, such as safety approvals or a different voltage. Compare entry modes: exporting through a distributor is cheap but slow, a partner shares the risk, buying a local company is fast but expensive. Build a simple business case with a break-even year. A good recommendation often names a first step that tests the market before the big bet, such as one region or one retailer.

A weak answer: "Yes, the US is a big market." Big is not the same as winnable. Follow-up: "The client has the money for only one market this decade." Compare the US with the next-best option on the same criteria, and say which you would choose and why.

Structure, hypotheses and frameworks

How do you structure a problem you have never seen before?

What they check: that you can break any question into parts, without a template to lean on.

Short answer: "I write the client's question at the top, then split it into a few parts that do not overlap and together cover everything, which consultants call MECE: mutually exclusive, collectively exhaustive. I keep splitting each part until I reach questions I can answer with data. That picture is an issue tree, and it tells me what to analyse first."

The full answer: A good split follows how the business works: profit into revenue and costs, customers into segments, a process into its steps. Two tests help: does any fact fit in two boxes (then they overlap), and is any cause missing (then it is not complete). Keep the top level to three or four parts. Then mark the branch most likely to hold the answer and start there; you do not need to analyse every branch to solve the problem.

A weak answer: a list of ten unrelated questions. Follow-up: "Your tree splits costs into staff, IT and other. Is that complete?" Yes, because other closes it, but it is weak: "other" is often the biggest box. Split by fixed and variable, or by function, instead.

What does hypothesis-driven work mean, and when can it go wrong?

What they check: that you test a likely answer early instead of collecting every fact first, and that you drop it when the facts say so.

Short answer: "It means stating early what I think the answer is, a hypothesis, and then looking for the facts that would prove or disprove it. It saves weeks, because I only gather the data that matters. It goes wrong when I fall in love with the hypothesis and read every fact as support, so I write down in advance what result would prove me wrong."

The full answer: In week one of a project, the team writes its best guesses at the answer and the analysis that would test each. Each analysis either supports the hypothesis, kills it, or changes it. Share the hypotheses with the client early: they often know at once which ones are wrong. The risk is confirmation bias, the habit of seeing only what fits. Guard against it with a test agreed in advance and a colleague asked to argue the other side.

A weak answer: "I never guess; I wait for the data." On a ten-week project that means the answer arrives in week nine. Follow-up: "The data disproves your hypothesis two days before the steering meeting." Tell the manager at once, present the new finding honestly, and say what it means for the recommendation.

When do you use a framework, and when do you put it aside?

What they check: that you use known frameworks as a starting point and adapt them to the case, instead of reciting them.

Short answer: "A framework is a starting point, not a checklist to recite. I use one when it fits the question, for example the four Ps of marketing for a pricing question, but I build the structure from the client's situation. Interviewers notice at once when a candidate forces a standard framework onto a case it does not fit."

The full answer: Frameworks such as Porter's five forces, which look at the pressure from competitors, buyers, suppliers, new entrants and substitutes, or a simple value chain, help you not to forget an area. But a case about a hospital's waiting times needs the patient's journey, not five forces. Say the structure in the client's words, name only the parts that matter here, and add what is special about this case, such as a regulation or a single large customer.

A weak answer: starting every case with the same memorised list. Follow-up: "Which part of your structure would you drop first if we had only ten minutes?" The part least likely to change the decision, and say why.

Slides, storylines and clients

How do you build a slide document that a client's board will read?

What they check: that you can turn analysis into a short story with the answer first.

Short answer: "I write the storyline before any slide: the answer, the three reasons that support it, and the facts behind each. That is the pyramid principle, answer at the top and support below. Each slide then has a headline that is a full sentence with the finding, one chart that proves it, and the source. Reading only the headlines should tell the whole story."

The full answer: Start with the question the board must decide. Draft the headlines on paper and test them with the manager before building charts; changing a sentence is cheaper than redoing ten slides. One message a slide; the detail goes to an appendix. Charts should make the point at a glance: the bar that matters highlighted, the rest in grey. End with the decision needed and the next steps with owners and dates.

A weak answer: headlines such as "Market overview" that say nothing. Follow-up: "The partner says the story is upside down." Ask which part should lead, often the impact in money, and rewrite the headlines before touching the charts.

A senior client disagrees with your findings in a meeting. What do you do?

What they check: calm client handling: you neither give in nor argue in front of the room.

Short answer: "I thank them and ask what they see differently; often they know a fact we missed. If their point is right, I say so and we adjust. If the data holds, I offer to go through it with them after the meeting, so nobody loses face. Before the next meeting I would test the findings with them one to one, so there are no surprises."

The full answer: Disagreement in a meeting is usually a sign the findings were not shared beforehand. Good consultants pre-wire, meaning they walk each key person through the findings before the meeting. In the room, separate the facts from the conclusion: agree on the facts first. Some objections are about the person's own area looking bad; acknowledge that and focus on what can be done. Tell your manager straight after, with what was said.

A weak answer: "I would explain the data again until they agree." Follow-up: "The client asks you to change a number in the report." Change it only if it is wrong. If it is right, explain why, and escalate to your manager if the pressure continues.

The client's data is incomplete and the deadline is in three days. How do you still give an answer?

What they check: judgement under time pressure: an answer that is good enough, with its limits said openly.

Short answer: "I would say what the answer depends on and work with the best data we have: a sample instead of the full set, an estimate from a comparable unit, or a quick survey of the people who know. I give a range rather than a single number, say how confident we are, and say what we would check with more time."

The full answer: Decide first whether the missing data could change the recommendation. If not, an estimate is fine and you say so. If it could, show both cases: with this assumption we recommend A, with that one B, and here is how to find out which holds. Clients accept honest ranges; they do not forgive a precise number that later proves wrong.

A weak answer: "I would ask for an extension." Follow-up: "Your estimate turns out to be off by a third." Explain what caused it, show whether the recommendation still holds, and update it openly.

Technology consulting: ERP, cloud and delivery

A client wants to replace its old ERP with SAP S/4HANA. How would you plan the project?

What they check: in an ERP case, that you see the project as a business change with a system in it, not only a software installation.

Short answer: "I would start with the processes the system must run, such as order to cash and procure to pay, and decide where the client adopts the standard and where it truly needs its own way. Then the plan: design, build, data migration, testing with real users, training and the go-live, each with an owner. The biggest risks are usually data quality and people who keep working the old way."

The full answer: An ERP is the system that runs a company's core processes: finance, purchasing, sales, stock. Choose the approach: a new start with clean processes, a technical move of the old system, or a mix. Keep to the standard where you can, because every custom change costs time at each later upgrade. Clean the data early: customers, suppliers and materials that are wrong in the old system will be wrong in the new one. Plan several test cycles with key users from the business, and a go-live weekend with a way back. For a finance module such as SAP FI/COThe finance part of SAP; asked for in many technology consulting ads for finance projects., the month-end close must work in the first month.

A weak answer: talking only about servers and licences. Follow-up: "The client wants every old report rebuilt." Ask which are really used, rebuild those, and replace the rest with standard reports or a dashboard.

How would you plan a client's move of its applications to the cloud?

What they check: in a cloud migration case, that you decide per application, with the cost and the risk in view.

Short answer: "I would list the applications, find what each depends on, and decide for each whether to move it as it is, adapt it, replace it with a cloud service, or retire it. I would move a few simple ones first to learn, set up the security and cost rules before the big waves, and track the running cost from day one."

The full answer: Start with why the client wants to move: lower cost, speed, or an ending data-centre contract; the reason decides the order. For each application, the common choices are rehost (move it unchanged), replatform (small changes), refactor (rebuild it for the cloud), replace or retire. Group them into waves by what they depend on. Agree the rules first: who may create what, where data may be stored, how costs are tagged. Compare clouds such as AWSAsked in 2% of consultant ads (as of 5 Nov 2026).One of the large clouds; cloud consultants plan which systems move there and in what order. and AzureAsked in 5% of consultant ads (as of 5 Nov 2026).Microsoft's cloud, common in companies that already run Microsoft software. on what the client already runs and its contracts. Cost surprises come from systems left running, so someone must own the bill.

A weak answer: "Move everything at once to save money." Follow-up: "After the move the bill is higher than the old data centre." Find the idle and oversized resources, switch off what nobody uses, and commit to the stable rest for a lower price.

Halfway through an implementation, the client keeps adding requirements. What do you do?

What they check: that you protect the scope, the date and the relationship at the same time.

Short answer: "I would log each new request, estimate its effect on time and cost, and take them to the steering group with a choice: add it and move the date or the budget, swap it for something of equal size, or plan it for a later release. The client decides, in writing."

The full answer: Scope creep, the slow growth of what a project must deliver, is the most common reason projects run late. A clear change process agreed at the start makes these talks routine instead of a fight. Many technology projects work in agileAsked in 3% of consultant ads (as of 5 Nov 2026).Deliver in short cycles with the client's team, showing working results early instead of one big launch. cycles, often with ScrumAsked in 3% of consultant ads (as of 5 Nov 2026).A common agile method with sprints and fixed roles, used on many technology consulting projects.: a fixed team and a fixed rhythm, where new wishes go into a list ranked by the client's product owner. Then the question is not whether, but what drops out.

A weak answer: "I say yes to keep the client happy." Follow-up: "The client says the request was always in scope." Go back to the signed scope and the minutes, calmly, and involve your manager if needed.

The fit interview

Why consulting, and why this firm?

What they check: a fit question tests whether your reasons are real and whether you know what the job asks of you.

Short answer: "I like solving a new problem every few months, with people who push my thinking, and I want to learn how many businesses work fast. Your firm's work in my sector and the way you put recommendations into practice with the client, which I heard about from people who work here, is why I applied to you."

The full answer: Give two or three reasons that come from your own experience: a project at university or at work where you enjoyed the problem solving, a time you worked with a consultant. Show you know the hard parts too: travel, long weeks in busy phases, and a client who can stop the project. For the firm, name something specific: its sector focus, its size, its approach, a conversation with someone who works there. Never only the brand.

A weak answer: "For the exit options and the money." Follow-up: "Where do you see yourself in five years?" Growing inside consulting, with a sector or a skill you want to master.

Tell me about a time you persuaded someone more senior who disagreed with you.

What they check: influence without authority, which consultants need every day with clients.

Short answer: "Use one real story in four parts: the situation, what I had to achieve, what I did, and the result. For example: I showed my manager that our biggest customer was not profitable. He disagreed. I rebuilt the numbers with him, line by line, so the conclusion became ours, and we renegotiated the contract."

The full answer: Interviewers listen for what you did, so say "I", not "we". Show that you listened first and understood why the person disagreed; often they had a fair point that changed your proposal. Show the facts you brought and how you made it easy for them to change their mind. End with the result and what you learned. Keep it under three minutes and prepare two more stories: a time you led a team, and a time you failed.

A weak answer: a story where you were simply right and the other person gave in. Follow-up: "What would you do differently?" Name one real thing, such as involving them earlier.

Senior consultant, manager, senior manager and partner

These questions come up for senior consultant, manager, senior manager, director and partner roles, where you run the project and then the client relationship.

As a manager, how do you run an engagement from the first week to the final presentation?

What they check: that you can deliver a project through a team, on time and within budget, with a happy client.

Short answer: "In week one I agree the questions, the scope and how success is measured with the client sponsor, write the first hypotheses, and give each consultant a clear workstream. I run a short team meeting each day, a weekly update with the client, and a steering meeting at each milestone. I review every slide before the client sees it, and I watch the hours against the budget each week."

The full answer: An engagement is the consulting word for a project with one client. The manager is the link between the partner, the team and the client: the team needs clear work and feedback, the client needs no surprises, the partner needs to know the risks early. Plan the storyline from the end: what must the final document say, and which analysis feeds each part. Protect the team from late-night rework by agreeing changes early. Coach each consultant on one skill during the project, and write honest feedback at the end.

A weak answer: "I do the hardest analysis myself." Then nobody manages the project. Follow-up: "Two weeks before the end, the sponsor changes the question." Agree what the new question means for scope and date, write it down, and involve the partner.

As a senior manager or partner, how do you win new work and grow a client relationship?

What they check: at the top of the ladder, that you can sell work by being useful, not by selling hard.

Short answer: "I build relationships before there is a project: I meet clients regularly, share ideas about their problems, and bring something useful each time. When a need appears, I write a proposal that restates their problem better than they could, with a clear approach, the team and a fair price. Most new work then comes from clients who were happy with the last project."

The full answer: At senior manager level you start to sell; directors and partners are measured on it. A proposal must show you understand the problem, set out the approach and the result, and name the people who will do the work. Agree the price and the scopeAsked in 3% of consultant ads (as of 5 Nov 2026).Agree scope, price and changes with a client without giving the project away. so that the project can succeed; a cheap project that fails costs the relationship. Grow an account by delivering well and noticing the next problem. Partners also build the firm: they train managers, shape a sector offer and lead peopleAsked in 3% of consultant ads (as of 5 Nov 2026).Lead a project team and coach junior consultants while the client watches the result..

A weak answer: "I would call my network and offer a discount." Follow-up: "The client asks you to start before the contract is signed." Agree a short written letter of intent first, so the work and the payment are clear.

Your CV for these questions

Every project on your CV becomes a question in the fit round. Lines like these give you a story ready for each; put each under the job where it happened:

Scoring your CV against a real consultant job shows which kinds of project that firm looks for, and where your CV gives no proof yet.