Strategy & advisory. The case interview: structure an ambiguous business problem, do quick maths, and drive to a recommendation.
A sample of the Consulting bank: the first 3 questions of every topic, with the model answers. 284 questions in total.
4 questions in this topic
Clarify the objective and success metric, take a moment to build a structured (MECE) framework tailored to the problem, walk the interviewer through it, then drive the analysis branch by branch, doing maths cleanly, stating insights as you go, and checking back to the question. End with a clear, prioritised recommendation, risks and next steps.
Four buckets: (1) the market, size, growth, segments, competition, profitability; (2) the client, capabilities, cost position, brand, right to win; (3) entry mode, build vs buy vs partner, and the economics/payback of each; (4) risks and the decision, does it beat the client's hurdle and alternatives? State a hypothesis early and drive to a go/no-go with conditions.
Strategic fit (does it advance the goal, capabilities, market access, scale), standalone value and the price/synergies (cost and revenue, net of integration cost and dis-synergies), and risks (integration, culture, customer overlap, regulatory). Then: is the value created greater than the premium paid, and is acquiring better than build/partner? Recommendation first, then the two or three swing factors.
1 more question on Case, with model answers, inside.
16 questions in this topic
Estimate top-down: population (~9m) × share who buy coffee daily × average cups, adjusted for commuters/tourists; sanity-check bottom-up (number of outlets × cups per outlet per day). State assumptions, keep the arithmetic clean and round sensibly. The number matters less than a defensible structure.
Bottom-up or top-down, pick one and be consistent. Top-down: ~33m cars, average fill frequency and litres per year → total fuel demand → ÷ throughput per station. Or per-capita: ~67m people, stations per 10k people from a known anchor. Aim for the right order of magnitude (the real figure is ~8,000), and sanity-check the two approaches against each other.
Top-down: population → cars → total miles driven → fuel consumed → per-station throughput. Or bottom-up from towns × stations per town. State assumptions, do the arithmetic transparently, and sanity-check the order of magnitude. The interviewer grades the structure and numeracy, not the exact figure.
13 more questions on Market sizing, with model answers, inside.
36 questions in this topic
48. 10% = 32, 5% = 16, sum = 48. Break percentages into easy 10%/5%/1% chunks and add, accuracy and speed both count.
30%. Change = 120 on a 400 base; 120/400 = 0.30. Break it down: 10% of 400 = 40, and 120/40 = 3, so 30%. Say the method out loud and keep the arithmetic clean, the interviewer is checking structure and accuracy under mild pressure.
96. 10% = 64, 5% = 32, sum = 96. Split percentages into 10/5/1% blocks: fast and error-proof under case pressure.
33 more questions on Maths, with model answers, inside.
26 questions in this topic
Variety of problems and industries, a steep learning curve, working with senior stakeholders on real decisions, and structured problem-solving. For "why this firm", reference something specific, a practice area, their approach, people you've met, not generic prestige.
Use a structured story with real depth: the situation and stakes, your specific role and the obstacle, the actions YOU took (aligning people, making the call, handling resistance), and a concrete, quantified result plus what you learned. Firms like McKinsey probe hard on personal impact, so pick an example you can go three or four "why/how" layers deep on.
Variety of problems and industries, a steep learning curve, working with smart teams on real strategic questions, and exposure to how businesses actually make decisions: a strong foundation whatever you do next. Ground it in a specific experience (a project, a case competition) rather than clichés, and tie it to why this firm.
23 more questions on Fit, with model answers, inside.
12 questions in this topic
Profit = revenue − cost. Split revenue (price × volume, by segment/product/geography) and cost (fixed vs variable, by function) to isolate where the change is, then whether it's internal (operations, mix, pricing) or external (market, competition, input costs). Localise the driver before recommending action.
Flat revenue with lower profit points at costs or mix. Split costs into fixed vs variable and by function to find what rose (input costs, labour, overhead), and check mix, are you selling more low-margin product at the same total revenue? Isolate the driver (internal vs external), quantify it, then target the fix. Always localise before recommending.
Trend-decompose the margin: is it gross margin (input costs, pricing power, mix shifting to lower-margin lines) or operating margin (fixed-cost growth, overhead creep, one-offs)? Compare each cost and revenue line as a % of revenue over the three years, benchmark against competitors, and isolate the biggest and most persistent driver. A gradual margin decline usually signals a structural issue (cost inflation, pricing erosion, or mix) rather than a one-off: target that.
9 more questions on Profitability, with model answers, inside.
9 questions in this topic
Take a moment, then lay out a MECE structure tailored to the question: don't force a textbook framework. Restate the objective, ask a clarifying question or two, break the problem into logical buckets (e.g. for profitability: revenue and costs; for entry: market, competition, capabilities, economics), state your hypothesis, and prioritise where to dig. Structure first, then analysis, then a recommendation.
Helpful as a mental checklist to ensure you're MECE and don't miss a dimension; harmful when applied mechanically to a problem it doesn't fit, producing generic, unprioritised analysis. Great candidates build a bespoke structure for the specific question and use frameworks as scaffolding, not a script. Tailor, don't recite.
Hypothesis-driven: form a view of the likely answer, identify which analysis would most confirm or kill it, and start there: the 20% of work that drives 80% of the insight. Sequence by impact and ease, and revisit as findings come in. Consultants don't boil the ocean; they test the critical assumption first.
6 more questions on Framework, with model answers, inside.
9 questions in this topic
Structure it: market attractiveness (size, growth, profitability, competition), the client's ability to win (capabilities, cost position, brand, distribution), the economics (investment, expected returns, breakeven, risk), and entry mode (build, buy or partner). Conclude with a clear recommendation and the key risk. Attractiveness and ability-to-win must both hold.
Assess market attractiveness by country (size, growth, competition, margins), the retailer's right to win (brand, format, cost position, supply chain reach), the economics (investment, breakeven, local cost structures), regulatory/labour differences, and entry mode (organic, acquisition, franchise, JV). Europe isn't one market: prioritise the best country and mode. Conclude with a phased recommendation and the key risk.
Market size and growth (broadband/mobile penetration, willingness to pay), competition (local incumbents, piracy), content strategy (local vs global, licensing costs), pricing and payment infrastructure, regulation, and the economics (subscriber acquisition cost, churn, content spend to breakeven). Content localisation and price sensitivity often make or break it. Recommend a mode and a phased plan.
6 more questions on Market Entry, with model answers, inside.
21 questions in this topic
Organic: existing customers (price, volume, cross-sell, churn reduction), new customers (segments, geographies, channels), new products/services; and inorganic: M&A/partnerships. Assess the size and feasibility of each lever, prioritise by impact and ease, and check the client can execute. Quantify the top two or three levers to hit the target.
Doubling in 5 years needs ~15% CAGR: demanding for a mature business. Assess the market's organic ceiling, then the levers: share gains, adjacencies/new products, geographic expansion, pricing, and M&A. Quantify each lever's contribution against the ~15% target; if organic can't get there, M&A must fill the gap. Be honest if the target is unrealistic and reframe.
Decompose growth into new logos vs expansion (net revenue retention) vs churn. Slowing usually means market saturation in the core segment, weakening land-and-expand, rising competition/churn, or a sales-efficiency problem (CAC up). Look at pipeline, win rates, retention cohorts and pricing. The fix depends on which broke: new markets/products, retention, or go-to-market efficiency.
18 more questions on Growth, with model answers, inside.
15 questions in this topic
Map the process end to end, measure throughput and capacity at each step, and find the constraint (the slowest station that gates the whole line). Then ask whether it's a capacity, quality/rework, downtime, or scheduling issue, and target that step: improving anything but the bottleneck doesn't raise output (theory of constraints).
Break down the cost base: transport/fuel, labour, warehousing, inventory carrying, and network design (routes, hub locations, load factors). Benchmark unit costs vs peers and history, and separate volume-driven from efficiency-driven increases. Target the biggest, most controllable lever: often network optimisation and load/route efficiency: with quick wins first.
Diagnose demand (call volume drivers: could some be deflected to self-service?) and capacity (staffing vs demand curve, handle time, first-call resolution, scheduling). Reduce avoidable contacts, match staffing to demand peaks, cut handle time via training/tools, and lift first-call resolution to reduce repeat calls. Attack both the demand and the efficiency side; measure the impact.
12 more questions on Operations, with model answers, inside.
11 questions in this topic
Externally: new or aggressive competitors, substitutes, shifting customer preferences, channel changes. Internally: price uncompetitiveness, product/quality gaps, weaker distribution or marketing, service issues. Segment where the share loss is concentrated (product, geography, customer type) to localise the cause, then test the most likely driver with data.
Compare total cost (in-house production cost + capex + overhead vs supplier price), plus strategic factors: control over quality and IP, supply security, flexibility, and whether it's a core capability. Make it if it's strategic or cheaper at scale; buy it if it's commoditised, capital-intensive, or better done by specialists. Quantify then weigh the qualitative.
Structure across the launch funnel: awareness (did target customers know?), trial (did they try: price, availability, positioning?), and repeat (did they come back: product, value, experience?). Layer in product-market fit, competition and channel. Localise where the funnel broke with data, then target the fix. Diagnose the stage before prescribing.
8 more questions on Business, with model answers, inside.
20 questions in this topic
Three lenses: cost-plus (a floor), competitor pricing (the reference), and value-based (what the customer's willingness-to-pay implies given the value delivered): the last usually captures the most. Consider positioning, price elasticity, segmentation and lifecycle. Value-based pricing anchored to the customer's economics beats cost-plus for anything differentiated.
Quantify the extra value delivered to the customer (cost saved, revenue enabled) and capture a share of it through value-based pricing: likely a premium: segmented by willingness-to-pay, while watching elasticity and competitive response. Test with a pilot. Leaving value on the table is a common, high-margin fix: price to value, not to cost or to the competitor.
Compute the breakeven volume increase: with a given contribution margin, a 10% price cut needs volume to rise enough to keep total contribution flat (the lower the margin, the more volume you need: often a lot). Then assess whether the market/elasticity can deliver that, and the competitive response (a price war). Usually price cuts require implausibly large volume gains: quantify before recommending.
17 more questions on Pricing, with model answers, inside.
8 questions in this topic
Same-store sales = traffic × conversion × average transaction value. Isolate which fell: fewer visits (competition, location, e-commerce shift, marketing), lower conversion (assortment, pricing, service, stockouts), or smaller baskets (mix, promotions). Compare to peers and channels to separate a company problem from a market/channel shift, then target the driver.
Test three things: is the new market attractive (size, growth, profitability, competition)? Can the company win there (transferable capabilities, brand, cost position, distribution)? And does it create value versus alternatives (better than returning cash, and better owned together than separately)? Diversification destroys value when there's no real synergy: the "better-off" test is key.
Compare on speed, cost, risk and control. Build: cheapest long-term and full control, but slow and risky if you lack the capability. Buy: fast and acquires proven capability, but expensive and carries integration risk. Partner: fast and low-capital, but shares value and cedes some control. Match the choice to how strategic/core the capability is and how fast you must move.
5 more questions on Strategy, with model answers, inside.
14 questions in this topic
Strategic rationale (market share, synergies, capability, defensiveness), target attractiveness (growth, margins, position), synergies quantified (cost and revenue, net of costs to achieve), valuation and price vs value, integration risk and the client's track record, and alternatives (build vs partner). Conclude with a clear yes/no, the price it's worth, and the biggest risk.
Diagnose which synergies (cost vs revenue) are missing and why: integration delays, culture/customer attrition, over-optimistic targets, or execution gaps. Re-baseline to a realistic number, prioritise the achievable cost synergies (usually more controllable), fix integration blockers, and manage stakeholder expectations. Revenue synergies commonly disappoint: focus effort where value is still recoverable.
Compare build vs buy on speed, cost, risk, and strategic control. Buying is faster and de-risks execution (proven team, product, customers) but costs a premium and carries integration risk; building is cheaper and gives full control but takes years and may let competitors move first. If the capability is core to differentiation and time is short, buy; if it's a commodity capability, build.
11 more questions on M&A, with model answers, inside.
10 questions in this topic
Build the business case: incremental revenue and cost savings the plant generates, the upfront and ongoing capex, then NPV/IRR against the hurdle rate, plus payback and sensitivity to key assumptions (demand, price, utilisation). Add strategic factors (capacity need, competitive response, flexibility). Recommend based on risk-adjusted NPV, not payback alone.
Payback ~3.3 years; NPV positive at reasonable discount rates (IRR ~15-20% on a 5-7 year life). Beyond the numbers, weigh reliability/quality gains, capacity/flexibility, implementation risk, and workforce impact. Recommend if the risk-adjusted NPV clears the hurdle and it fits strategy, but sensitivity-test the savings and adoption assumptions before committing.
Compare risk-adjusted returns: paying down debt is safest if leverage is a concern or debt is expensive; buybacks make sense if the stock is undervalued and there's no better internal use; growth investment wins if the client has projects earning above its cost of capital. Prioritise by return versus risk, not just what's available.
7 more questions on Investment, with model answers, inside.
20 questions in this topic
Stabilise cash first (13-week cash flow, cut discretionary spend, manage working capital and creditors), then diagnose the core problem (revenue vs cost vs one-off), fix the value drivers, and set a viable plan/structure. Cash before strategy: a business that runs out of cash doesn't get to execute the clever turnaround. Triage, then transform.
Whether the business is fundamentally viable: is the core product and market still attractive, or is this a structural decline? A turnaround only makes sense if underlying demand and unit economics can work; if the business model itself is broken, the answer is exit or wind-down, not a turnaround plan.
Stabilise (stop the cash bleed: cost cuts, working capital, covenant fixes), fix the core (pricing, portfolio, operational performance in the business that will survive), and grow again once stable. Sequencing matters, you can't grow your way out of a cash crisis.
17 more questions on Turnaround, with model answers, inside.
8 questions in this topic
Assess where value is shifting (which customers/categories move online), the client's assets (stores, brand, data, supply chain) and how to leverage them (omnichannel, click-and-collect, fulfilment from store), the economics of online vs offline, and where to invest vs defend vs exit. The answer is rarely "go all-online": it's a differentiated, economics-led response using existing strengths.
Concretely: which processes and channels are moving from manual or physical to digital or data-driven, e-commerce, digital marketing, automated operations, and what capability, technology, talent, ways of working, gets built to sustain it. A good answer names specific business processes being transformed, not just "going digital."
Rank by value at stake, cost saved or revenue enabled, against ease of implementation, data readiness, system complexity, change management need. Start with high-value, high-feasibility quick wins to build momentum and fund the harder, higher-value but more complex initiatives later.
5 more questions on Digital, with model answers, inside.
17 questions in this topic
Segment costs (fixed/variable, by function, value-adding vs not), benchmark vs peers to find the fat, and target non-customer-facing overhead, procurement, process inefficiency and footprint before front-line/growth investments. Use zero-based thinking, protect the capabilities that drive revenue, and sequence quick wins then structural change. Cut with a scalpel, not an axe.
Segment the cost base (COGS: materials, labour, overhead; SG&A) and benchmark vs peers to find the fat. Attack procurement, footprint/overhead, process inefficiency and low-value SKUs first; protect R&D, quality and customer-facing capabilities that drive revenue. Use zero-based budgeting for discretionary spend, sequence quick wins then structural change, and quantify each lever. Cut deep but not into the muscle.
Cost-cutting is typically a one-off, often across-the-board reduction that can erode capability if applied blindly; a cost transformation redesigns how work gets done to permanently lower the cost base without cutting into what drives value. Boards should ask for the latter even when they initially request the former.
14 more questions on Cost, with model answers, inside.
8 questions in this topic
Profit = net interest income + fees − costs − provisions. Attack each: net interest margin (loan pricing, funding mix, deposit costs), fee income (products, cross-sell), cost/income ratio (branches, digitisation, headcount), and credit losses (underwriting, collections). Benchmark against peers to find the gap, and weigh regulatory/capital constraints. Prioritise the biggest, most feasible lever.
Map the patient flow end to end, measure throughput and delay at each stage, and find the bottleneck (beds, staff, diagnostics, discharge). Separate demand issues (volume, seasonality) from capacity and process issues (scheduling, handoffs, discharge delays). Target the constraint, pilot a fix, and measure. Operations logic: theory of constraints: applied to healthcare.
Segment the churn (which customers, plans, regions) and diagnose drivers: price/value vs the rival, network quality/coverage, customer service, contract friction, and switching incentives. Compare the offer and experience head-to-head, quantify the churn's revenue impact, and prioritise retention levers (pricing, network investment, loyalty, service). Localise the leak before spending to plug it.
5 more questions on Industry, with model answers, inside.
20 questions in this topic
Diagnose the barriers: incentives (siloed targets/comp), structure (no shared accountability), systems (no shared customer data), and culture. Fix incentives to reward cross-sell, create shared goals/governance, enable with common data/tools, and get visible leadership sponsorship. Structural and incentive fixes beat exhortation: people follow how they're measured and paid.
Centralise when scale economies matter most, buying power, standardisation, shared expertise, and needs are similar across units; decentralise when speed and local responsiveness matter most. Many companies land on a hybrid: centralise the transactional, scale-driven parts and decentralise the judgment calls close to the business.
Who has the authority to make which decisions, and who just needs to be consulted or informed. A clean org chart with muddled decision rights still produces slow, political decision-making; clarifying decision rights often fixes more dysfunction than redrawing boxes.
17 more questions on Organisation, with model answers, inside.