Roland Berger Case Interview Examples: 8 Candidate-Led Practice Cases (2026)

Roland Berger case interviews are candidate-led and industrially flavoured: you propose the structure, request each piece of data, and drive to a quantified recommendation, usually inside an automotive, industrial, energy or public-infrastructure setting. Founded in Munich in 1967 and still entirely partner-owned, the firm passed one billion euros of revenue in 2023 and runs around 3,500 people across more than 50 offices. Its restructuring and performance transformation practice is the reason its cases skew toward turnarounds and cost programmes far more than a generic strategy case set does. The firm publishes two complete case webinars of its own, Transit-Oriented Development and 3D-Printed Hip Implants, each in two parts.

This page gives you 8 worked practice cases in those archetypes, decodes the two official webinar cases, and shows exactly which skill each one is testing. For the process itself, the round structure, the online test and the group case mechanics, use the Roland Berger case interview guide.

What should a Roland Berger practice case show?

A useful Roland Berger example shows five things: the client decision, the constraint that makes it hard, the first structure you would propose, the specific evidence you would request, and a recommendation with a number attached. A case that stops at "here is the framework" has skipped the part the interviewer is actually scoring.

Work each case in that order: state the decision, build the tree, request the highest-value evidence, then close. Reading a worked solution top to bottom teaches you the pattern; covering the answer and rebuilding it is what actually moves your score.

How Roland Berger cases differ from McKinsey, BCG and Bain

The format similarities are real, and so are the differences. Roland Berger is candidate-led like BCG and Bain, but the content bias is much stronger, and the recommendation culture is noticeably more decisive.

What this means for practice. If you have only rehearsed interviewer-led cases, the hardest adjustment is not the content, it is the silence. Nobody hands you the next exhibit. You have to say "I would like to see the cost breakdown by plant for the last three years" and then do something with it. And if you have only rehearsed consumer and tech prompts, the industrial numbers will slow you down: capacity utilisation, tonnes, fixed cost absorption, per-unit contribution.

The fastest way to feel the difference is to drive one European profitability reset yourself, proposing the structure and calling for each exhibit rather than waiting.

The five question types inside a Roland Berger case

A Roland Berger case is a sequence of separately scored moves. The firm names analytical skill, conceptual ability and communication as the three competencies it evaluates, and each of those shows up in a specific ask. Rehearsing the moves separately is faster than running full cases you cannot diagnose afterwards.

The one you cannot grade yourself is the first. Almost every candidate believes their structure was fine, because they recognise it. Get one scored before you read the examples below, so you know whether your tree is genuinely hypothesis-led or a template you retrieved.

Restructuring and turnaround: the archetype to own

If you prepare one archetype deeply for Roland Berger, make it this one. Restructuring and performance transformation is a named practice at the firm, not a niche, and its published research is built on it. In the study Roland Berger ran with Civey in June 2024, 500 C-level executives were surveyed: 73 percent said their company was threatened by structural upheaval, 42 percent called that threat serious, and 56 percent said conventional cost-driven restructuring is no longer sufficient. The stated priorities were digitalisation and AI at 40 percent, conventional cost cutting at 34 percent, entering new markets at 33 percent, and changing the business model at 27 percent.

That last set of numbers is the tell for how to answer. A pure cost answer is the answer the firm's own clients have already outgrown. The structure that scores here has three peer branches and an explicit sequence:

1. Stabilise. Cash first: working capital, payment terms, capex freeze, the loss-making contracts you can exit inside a quarter. This branch is measured in weeks.
2. Restructure the cost base. Footprint, headcount, procurement, product complexity. Measured in quarters, and constrained by works councils and notice periods in most European jurisdictions.
3. Rebuild the earnings model. Pricing, mix, portfolio pruning, new revenue. Measured in years, and this is the branch weak candidates skip entirely.

Then say which one you would test first and why. In a cash crisis the answer is almost always branch one, because the other two need runway to work. Saying that out loud, with the reason, is the conceptual-ability signal the firm says it is looking for. For the full method, work through the restructuring case interview guide and the cost reduction case patterns.

The official Roland Berger webinar cases, decoded

Roland Berger publishes two full case walkthroughs on its careers site and on YouTube, each split into two parts and available in English and German. They are the highest-value material available to you, because you see how an actual interviewer reacts in real time, where they interrupt, and what they let slide.

Case one, Transit-Oriented Development. A transit operator profitability and revenue problem. The interesting move is that the revenue in question is not only fare revenue: transit-oriented development means the land and property around stations is part of the earnings model. Candidates who structure this as "fares times riders minus operating cost" and stop there miss the entire second half of the case. Watch for how the interviewer nudges toward the non-fare revenue branch and how long the candidate takes to get there.

Case two, 3D-Printed Hip Implants. An additive-manufacturing market entry question. Two things make it harder than a standard entry case: the addressable market has to be built from a medical population rather than a consumer one, and the economics of 3D printing invert the usual assumption that unit cost falls with volume. Watch for how the sizing is assembled from procedure counts rather than from a top-down market figure.

Working these two well is worth more than a dozen generic cases, for a reason that is easy to miss: they are the only cases where you can check your instinct against how a Roland Berger interviewer actually behaves. Take notes on the interviewer's questions, not just the candidate's answers.

Both cases lean hard on the same underlying skill, which is building a defensible number from drivers when nobody gives you a market size. That is the skill to get graded before you attempt case four below.

8 Roland Berger case examples with worked solutions

Every number inside these cases is a case assumption written for practice, not a published figure about a real company. Use them the way an interviewer would: state the structure, ask for the evidence, and only then read the worked path.

Case 1: Restructuring, German automotive supplier margin collapse

Prompt: A German tier-one automotive supplier making powertrain components has gone from an 8 percent EBIT margin to minus 2 percent in three years. Revenue is flat at 900 million euros. The bank syndicate has asked for a plan within six weeks. What do you do?

How to drive it: Do not open with a profitability tree. Open by asking about the constraint: how much cash runway is there, and what covenant is breaking? Then propose the stabilise, restructure, rebuild sequence and say you would test stabilise first. Ask for the cash flow forecast, the plant-level contribution margin, and the revenue split between combustion and electric platforms.

Key insight pattern: In a powertrain supplier, the margin collapse is almost never operational inefficiency alone. It is a demand-mix shift: combustion volumes are declining, the fixed asset base was sized for them, and the electric programmes are early, low-volume and priced on launch terms. That is a footprint and portfolio problem wearing a cost problem's clothes.

Worked approach:

1. Quantify the fixed cost absorption gap. If combustion volumes fell 30 percent and combustion carries 70 percent of revenue, roughly 21 percent of the volume that used to absorb the fixed base has gone. At a 35 percent contribution margin on 630 million euros of combustion revenue, that is about 66 million euros of lost contribution against a broadly unchanged fixed base, which more than explains a 10-point margin swing on 900 million euros.
2. Test whether the electric programmes are dilutive or merely young. If electric revenue is 90 million euros at a 12 percent contribution margin against a 35 percent combustion margin, the mix shift alone drags blended contribution down even at constant volume. If the launch pricing steps up at a defined volume threshold, the programmes are young. If it does not, they are dilutive and were mispriced.
3. Sequence the plan. Weeks one to six: cash. Stretch supplier terms, freeze non-programme capex, exit or reprice the two lowest-contribution combustion contracts. Quarters one to four: footprint. Consolidate combustion machining into the two highest-utilisation plants and convert one to electric assembly rather than closing it, because retained capacity is cheaper than rebuilt capacity. Years one to three: portfolio. Reprice the electric book at renegotiation points and exit the sub-scale product families.
4. Recommendation: a three-horizon plan that returns roughly 4 points of margin from footprint consolidation and roughly 3 points from repricing and portfolio exits, with the cash actions buying the runway to execute. Principal risk is works council timelines on the consolidation, so start the consultation in week one rather than after the plan is approved.

The blended-margin and absorption arithmetic in step one is where most candidates lose the room. The numbers are not hard, but the weighting is, and there is no partial credit for getting the direction right and the magnitude wrong. Rep that exact calculation shape once and see where the feedback lands.

Case 2: Cost reduction, European retail chain under a 30 percent target

Prompt: A high street retail chain wants to cut its cost base by 30 percent following a demand shock. How would you approach it? This prompt circulates widely in Roland Berger candidate reports and is worth rehearsing verbatim.

How to drive it: The trap is to start listing cost lines. Start instead by asking what 30 percent is measured against, and over what period. Thirty percent of total cost including cost of goods sold is a different exercise from 30 percent of operating expense. Then propose three branches: cost of goods, store network and fixed operating cost, and central overhead, and say you would test the store network first because it is the largest addressable block in retail.

Key insight pattern: In a store-based retailer, roughly 20 to 25 percent of stores usually generate a small or negative four-wall contribution. The right first move is not a uniform cut, it is finding the tail and deciding which part of it is fixable versus closable. A uniform 30 percent cut applied across a healthy estate destroys the stores that were funding the recovery.

Worked approach:

• Segment the estate by four-wall contribution. If 900 stores generate 1.8 billion euros of revenue and the bottom 180 stores generate 200 million euros of revenue at a negative 2 percent four-wall contribution, closing them removes about 4 million euros of losses plus their share of fixed overhead, but also removes 200 million euros of revenue and any online demand they anchor.
• Split the tail into fixable and structural. Fixable means a lease renegotiation or a labour schedule change flips the store positive. Structural means the catchment has gone. Only the structural half should close.
• Attack cost of goods through complexity, not price. Cutting stock-keeping units by a third usually raises the volume per remaining line enough to reopen supplier terms, which is a durable saving, where a one-off price demand is not.
• Recommendation: close the structurally negative stores, renegotiate the fixable tail, cut range complexity, and take central overhead last because it is the smallest block and the most damaging to cut first. State plainly that 30 percent measured on total cost is likely unreachable inside one year without impairing the recovery, and give the number you can reach and by when. Being willing to challenge the target is the answer that separates candidates here.

For the full method behind this shape, work the cost reduction case interview patterns.

Case 3: Transit revenue, regional rail operator

Prompt: A regional rail operator carries 40 million passenger journeys a year and is losing 25 million euros annually before subsidy. The transport authority will not raise the subsidy. Ridership is flat. What levers exist?

How to drive it: Ask for the revenue split first, and specifically whether any non-fare revenue exists today. This mirrors the structure of Roland Berger's own transit case. Propose branches for fare revenue, non-fare revenue, and operating cost, and say you would test non-fare first precisely because ridership is flat and fares are politically constrained.

Key insight pattern: A transit operator is also a property and footfall business. Stations aggregate people daily, which is a commercially valuable asset that most operators monetise poorly. The candidate who only optimises fares and rosters is solving half the case.

Worked approach:

• Fare revenue: at 40 million journeys and an average fare of 3.20 euros, fare revenue is 128 million euros. A 5 percent yield improvement through off-peak pricing and fare-evasion enforcement is worth about 6.4 million euros, assuming evasion runs near the sector-typical mid single digits. Note that raising the headline fare is usually blocked by the authority, so yield has to come from mix and enforcement.
• Non-fare revenue: retail concessions, advertising, parking and, at the larger stations, development rights on adjacent land. If 12 major stations each support 400,000 euros of incremental concession and advertising income, that is 4.8 million euros with limited capital.
• Operating cost: crew rostering and energy are the two large controllable lines. A 4 percent reduction on a 110 million euro operating base is 4.4 million euros.
• Recommendation: the three levers together close roughly 15.6 million euros of a 25 million euro gap, so say clearly that the gap does not close on operating levers alone. The remainder requires either a service pattern change on the lowest-utilisation routes or a development partnership on station land. Name which one you would test first and what evidence would decide it.

Case 4: Market entry, additive manufacturing for orthopaedic implants

Prompt: A European medical devices manufacturer is deciding whether to build a 3D-printing capability for hip implants. Is this an attractive business? This mirrors Roland Berger's second published case.

How to drive it: Build the market from procedures, not from a top-down device market figure. Ask for the annual hip replacement volume in the target geography, the share of procedures where a custom implant is clinically indicated, and the price premium a custom implant commands. Then handle the economics separately, because additive manufacturing does not behave like injection moulding.

Key insight pattern: With 3D printing, unit cost is close to flat with volume. There is no tooling amortisation to spread, so the classic "scale it and the cost falls" reflex is wrong. The economics come from the price premium on customisation and from inventory that no longer has to be held in twelve sizes, not from volume.

Worked approach (bottom-up sizing):

• Take a market of roughly 250,000 hip replacements a year across the target countries as your working assumption, and state it as an assumption.
• Share clinically suited to a custom implant, primarily revision surgeries and anatomically atypical cases: assume 15 percent, giving about 37,500 procedures.
• Realistic penetration for a new entrant in five years against established players: assume 12 percent, giving about 4,500 units a year.
• Price: a standard implant at roughly 2,000 euros against a custom printed implant at roughly 3,200 euros, so 4,500 units gives about 14.4 million euros of revenue.
• Cost: printing plus post-processing plus regulatory and validation overhead at roughly 1,900 euros per unit gives about 8.6 million euros of gross profit, against a capital requirement in the tens of millions for printers, a validated clean facility and the regulatory pathway.
• Recommendation: attractive on margin, marginal on scale. The decision hinges on whether the capability is a platform for other implant categories or a single-product bet. Recommend entry only if the same validated facility can carry knee and spinal programmes, because the fixed regulatory and facility cost is the real barrier and it is shared. Principal risk is regulatory timeline, so the first test is the approval pathway duration, not the market size.

The structural move that made this case answerable was refusing the top-down market figure and building from procedures. If your driver chain wobbled there, that is the thing to fix before the next case, not the arithmetic.

Case 5: Industrial market entry, machinery maker entering India

Prompt: A German mid-sized maker of packaging machinery, 400 million euros of revenue, is considering entering India. Its machines sell for 600,000 euros in Europe. Should it enter, and how?

How to drive it: Structure around attractiveness, right to win, and entry mode, then say you would test right to win first, because a premium European machinery maker's usual failure in emerging markets is not demand, it is price positioning. Ask for the local competitive price points and the installed base by machine tier.

Key insight pattern: The European machine is engineered for labour costs and uptime economics that do not hold in the target market. A customer paying a fraction of European labour rates does not value automation the same way. The entry decision is really a product decision: enter with the European machine, or with a de-specified variant.

Worked approach:

• Attractiveness: assume the local packaging machinery market is around 900 million euros growing at 9 percent, split roughly 70 percent low-tier local machines at 120,000 to 200,000 euros and 30 percent imported mid and high tier.
• Right to win: at 600,000 euros the machine competes only in the top slice, perhaps 150 million euros of addressable value, against established international competitors already present. A de-specified variant at 280,000 euros opens a much larger band but risks cannibalising European sales and diluting the brand.
• Entry mode: greenfield sales subsidiary gives control and margin but takes three to four years to build a service network, which is decisive for machinery because uptime is the purchase criterion. Distributor gives speed and reach but weak service quality. A joint venture with a local manufacturer gives cost position and service reach at the cost of intellectual property exposure.
• Recommendation: enter, but with a de-specified variant manufactured locally through a joint venture, and ring-fence the premium machine for export customers only. The number that decides it is the service network build cost against the three-year revenue ramp, so request that comparison as your next test.

Case 6: Operations, chemicals footprint consolidation

Prompt: A specialty chemicals producer runs seven European plants at an average 62 percent capacity utilisation. Group EBIT margin is 6 percent against a peer benchmark of 12 percent. Should it consolidate the footprint?

How to drive it: Ask for utilisation, contribution margin and fixed cost by plant, plus the product-to-plant mapping and freight cost to major customers. The mapping matters more than candidates expect: a plant can look redundant on capacity and be irreplaceable on product qualification.

Key insight pattern: In chemicals, closing a plant is rarely a pure capacity arithmetic problem. Products are qualified at specific sites with specific customers, and requalification can take twelve to eighteen months. The consolidation candidate is the plant whose products can be requalified elsewhere fastest, not simply the emptiest one.

Worked approach:

• Capacity arithmetic: seven plants at 62 percent utilisation means the same volume fits into roughly five plants at 87 percent, which is a workable operating level with maintenance headroom.
• Fixed cost: if each plant carries roughly 14 million euros of fixed cost, closing two removes about 28 million euros gross, less severance, decommissioning and site remediation. On revenue of 700 million euros that is roughly 4 points of margin before one-offs, which closes two thirds of the peer gap.
• Constraint check: freight. Consolidating into two sites raises average delivery distance, and specialty chemicals freight is not trivial. Assume an incremental 4 million euros of annual logistics cost, which reduces the net benefit to roughly 24 million euros.
• Requalification: rank the two closure candidates by the number of customer-qualified products they hold. The right answer may be to close the second-emptiest plant rather than the emptiest, if the emptiest holds the qualifications.
• Recommendation: consolidate two plants, chosen on requalification exposure rather than utilisation alone, phased over eighteen months to run requalification in parallel with the ramp. Net benefit around 24 million euros a year, principal risk is a customer switching supplier during requalification, so pre-agree the transition with the top ten accounts before announcing.

Case 7: Energy transition, utility deciding on grid-scale storage

Prompt: A regional European utility is considering investing 200 million euros in grid-scale battery storage. Its board wants to know whether this is a strategic necessity or a distraction. What is your recommendation?

How to drive it: Separate the two questions the board actually asked. One is a returns question, and one is a positioning question, and they can have different answers. Propose branches for revenue stack, cost and technology risk, and strategic fit with the existing generation and network business. Ask for the price volatility data first, because storage revenue is a volatility play.

Key insight pattern: Battery storage does not earn from a single revenue line. It stacks: wholesale arbitrage, frequency response and capacity payments. Candidates who model only arbitrage will conclude the investment fails, because arbitrage alone usually does not clear the hurdle rate.

Worked approach:

• Revenue stack: assume a 200 megawatt-hour system. Arbitrage on a daily spread, frequency response contracted annually, and capacity market payments. Model each separately and state that the stack, not any single line, determines viability.
• Cannibalisation: as more storage enters the market the daily spread narrows, so the case has to assume a declining arbitrage margin over the asset life rather than today's spread held flat. Candidates who miss this overstate returns badly.
• Technology risk: cell degradation reduces usable capacity over time, and replacement economics land inside the asset life. Ask when the degradation curve crosses the contracted capacity commitment.
• Strategic fit: a utility with existing renewable generation gains a real advantage, because storage firms up its own intermittent output and reduces its imbalance costs. That is a benefit a pure-play storage investor cannot capture, and it belongs in the case.
• Recommendation: invest, but sized to the utility's own renewable portfolio rather than as a merchant asset, and contract the frequency response revenue before committing capital. The decisive next test is the imbalance cost saving on the existing portfolio, because that is the part of the return only this owner can earn. For more on this sector's case patterns, see the energy case interview guide.

Case 8: Due diligence, private equity carve-out of an industrial components division

Prompt: A private equity client is evaluating the carve-out of an industrial components division from a listed group. Revenue is 320 million euros, reported EBITDA is 38 million euros, and the asking price is 340 million euros. Is the deal attractive?

How to drive it: Say immediately that reported EBITDA on a carve-out is not the number you are buying, and ask for the standalone cost estimate. Then structure around market attractiveness, standalone earnings quality, and value creation levers. This is the archetype where candidates most often accept the given number and lose the case in the first minute.

Key insight pattern: A carve-out carries stranded costs. The division consumed group IT, finance, HR and procurement scale that disappear on day one. Standalone EBITDA is systematically lower than reported divisional EBITDA, and the gap is the whole deal.

Worked approach:

• Standalone adjustment: assume dis-synergies of 6 million euros a year across IT, finance, procurement scale loss and a standalone management team. Adjusted EBITDA is therefore 32 million euros, and the multiple moves from 8.9 times to 10.6 times, which is expensive for industrial components.
• Earnings quality: check customer concentration and contract length. If the top three customers are 45 percent of revenue on annually renewing contracts, that is a materially riskier 32 million euros than the same figure spread across fifty accounts.
• Market: is the underlying component demand tied to a declining platform? In industrials this is the equivalent of the combustion exposure in Case 1, and it should be tested explicitly rather than assumed stable.
• Value creation: pricing discipline, procurement re-tendering at standalone scale, and footprint. Quantify the two largest and state the entry multiple you would need for the deal to clear the fund's return threshold.
• Recommendation: proceed only below roughly 290 million euros, which restores a 9 times multiple on adjusted EBITDA and leaves room for the standalone build cost. Principal risk is customer concentration, so the first confirmatory test is contract renewal terms with the top three accounts.

The closing 60 seconds is where all eight of these cases are won or lost, and it is the part almost nobody rehearses in isolation. Practise compressing an analysis into a decision, a number, a risk and a next step.

The group case: how to practise something you cannot practise alone

Where a group exercise applies, candidate reports describe a consistent shape: roughly three to six candidates, a shared package of charts and financials, a short individual reading window of ten to fifteen minutes, a group discussion of fifteen to twenty minutes, and a joint presentation followed by interviewer questions. Assessment is on collaboration, listening and the quality of your contribution.

The failure modes are predictable and worth naming so you can avoid them:

• Competing rather than contributing. More than one candidate in a group can be hired. Talking over people costs you the collaboration score and gains you nothing.
• Going silent because someone else took the structure. Silence is scored as absence. If the structure is taken, take the quantification, the risk register or the synthesis.
• Never converging. Groups that debate the framework for fifteen minutes present nothing. Somebody has to say "we have eight minutes, let us lock the structure and split the analysis." That person usually scores well.
• Presenting four disconnected pieces. Assign one person to stitch the presentation into a single argument before you present.

Prepare it the way it is run: get three to five peers, take a real case package, and rehearse the exact timings. The group case interview guide has the full mechanics, and the consulting group case and inbox playbook covers the assessment-centre variants some offices use alongside it.

Self-diagnostic: are you practising like a Roland Berger candidate?

After working the eight cases above, answer these honestly.

1. Did you name the constraint before you built the tree? In cases 1, 2 and 6 the constraint (cash runway, target period, requalification time) determines the answer. A candidate who builds a beautiful profitability tree without asking what is binding has skipped the diagnosis.
2. Did you challenge a given number? Case 8 hands you an EBITDA figure that is wrong for the transaction, and case 2 hands you a target that may be unreachable. Accepting both is the most common failure at this firm, and challenging them politely is the fastest way to look like a consultant rather than a candidate.
3. Was your recommendation sequenced? Restructuring answers need a horizon: what happens in weeks, in quarters, in years. A flat list of ideas reads as a brainstorm.
4. Did your math carry units and periods? Industrial cases are full of tonnes, units, megawatt-hours and annualised figures. Losing the unit is how a correct method produces a wrong number.
5. Could you defend the sector logic? If you cannot say why powertrain volumes are declining or why storage revenue stacks, the interviewer will find the edge of your knowledge quickly. Reading around automotive cases and the top automotive consulting firms landscape closes most of that gap.

If more than two of those landed badly, the gap is method rather than reps. Learning Mode walks the case-solving sequence end to end before you spend another hour on full cases.

Roland Berger's official practice resources

Work the firm's own material before anything else. It is free, it is accurate about what the firm wants, and it is short.

• The two case webinars. Transit-Oriented Development and 3D-Printed Hip Implants, each in two parts, in English and German, linked from Roland Berger's case study preparation page and available on YouTube.
• The firm's own case guidance. The same page states plainly what it evaluates: structure your thinking, make realistic assumptions with common sense and basic data, separate critical from peripheral information, and treat the case as a dialogue. It also says the interviewers are not trying to trip you up, which is the cue to ask clarifying questions rather than guess.
• Published industry research. The Automotive Disruption Radar, the energy transition series and the restructuring studies on Roland Berger's publications hub are the fastest way to load the sector vocabulary the cases assume. Read the executive summary of two reports in the practice area named in your role.
• The reasoning test material. Where an online test applies, its vendor and rules come from your invitation. If it is a numerical and verbal battery, the numerical reasoning test guide covers the question types.

Once the firm's own material is exhausted, the constraint stops being content and becomes reps under pressure. Browse the case library and pick candidate-led profitability, market entry and operations prompts to run with the same loop.

How to use these Roland Berger practice cases

Run each case in the same sequence, or you are reading rather than practising.

Execution checklist

• ✓

  Read only the prompt, then cover the rest. The worked path is the answer key. Reading it first destroys the rep. Set a timer for two minutes and write your own structure before you scroll.

• ✓

  Name the constraint out loud. Cash runway, target period, regulatory timeline, requalification window. In six of the eight cases above the constraint decides the answer, and stating it first is what separates a diagnosis from a framework.

• ✓

  List the three exhibits you would request. Candidate-led means nothing arrives unasked. Write the exact exhibits, then check them against what the worked path actually uses. If you asked for data the answer never needed, your tree was not hypothesis-led.

• ✓

  Do the math on paper with units attached. Industrial cases run on tonnes, units and annualised figures. If the arithmetic slipped, rep it on the free math drills rather than moving to the next case.

• ✓

  Write the recommendation in four parts. Decision, number, principal risk, next test. Then compare against the worked path's closing paragraph and mark which of the four you left out.

• ✓

  Wire each weakness to one rep, not another case. Structure felt template-shaped, run free structure drills. Sizing wobbled, work the market sizing method. Exhibits took too long, work reading charts and exhibits.

• ✓

  Rehearse the group format with peers. Book three to five people, take one case package, and run the real timings from the group case section above. It is the only component you genuinely cannot rehearse alone.

• ✓

  Fix the application before the round, not after. Run the CV through the consulting resume grader and rebuild the ownership stories using the Roland Berger behavioral questions before the first interview, because neither is fixable once the process starts.

If you have time for only three of the eight, do cases 1, 3 and 8 back to back: one restructuring, one revenue-model problem, and one due diligence with a number you have to challenge. That trio covers most of what the firm actually asks. For worked examples across other firm formats, see case interview examples, and for the process, rounds and online test detail, the Roland Berger case interview guide carries the full walkthrough plus a worked automotive case.

Sources and Further Reading (checked July 31, 2026)

• Roland Berger case study preparation, official guidance and the two case webinars: rolandberger.com/en/Join/Your-application/Hints-tips
• Roland Berger restructuring survey of 500 C-level executives, June 2024: rolandberger.com/en/Insights/Publications/Why-restructuring-has-to-be-rethought.html
• Roland Berger Restructuring and Performance Transformation practice: rolandberger.com/en/Expertise/Services/Restructuring-Performance-Transformation
• Roland Berger insights and publications hub: rolandberger.com/en/Insights
• Roland Berger company overview, founding, ownership and scale: en.wikipedia.org/wiki/Roland_Berger_(company)
• PrepLounge Roland Berger interview guide and candidate reports: preplounge.com/en/blog/consulting/firms/roland-berger
• CaseBasix Roland Berger case interview guide, round structure and group case timings: casebasix.com/pages/roland-berger-case-interview-guide
• Roland Berger case webinar, Transit-Oriented Development part one: youtube.com/watch?v=uw0ZTxvBzkI