McKinsey Case Interview Examples: 20 Practice Cases (2026)

What McKinsey Actually Tests (The 4-Dimensional Rubric)

Every McKinsey case is scored on four dimensions. If you understand the rubric, you stop optimizing for the wrong things. Most candidates over-index on frameworks and under-invest in synthesis, which is where 4s and 5s separate from 3s.

• ●Structured thinking: Can you break a messy business problem into MECE parts in under 90 seconds?
• ●Quantitative reasoning: Can you drive a number to within 5% of a reasonable answer while narrating your logic?
• ●Business judgment: Can you identify which driver actually moves the business, not just list all of them?
• ●Synthesis: Can you land a single recommendation with one risk and one next step in 60 seconds?

The 3 Case Archetypes You Will Face

McKinsey cases fall into three archetypes. Every question you get in 2026 is a variant of one of these, or a hybrid. Knowing the archetype before you start structuring saves you 30 seconds and keeps your tree clean.

1. Profitability Cases (~40% of first rounds)

A client's profits are down. Diagnose why and recommend a fix. The universal profitability tree is Revenue minus Cost, then Revenue splits into Price times Volume, and Volume splits into Customers times Purchases per customer. Cost splits into Fixed and Variable. 95% of profitability cases are solved by picking the correct leaf of this tree and going deep.

2. Market Sizing Cases (~25% of first rounds)

Estimate a number from the ground up. The two valid paths are top-down (start from population, apply penetration) and bottom-up (start from one unit, scale up). Always state which you chose and why. Never mix paths mid-case. This is the single most common error McKinsey flags.

3. Market Entry / Strategy Cases (~35% of final rounds)

Should client X enter market Y? Frame this as four questions: Is the market attractive? Can we win? Is it strategically a fit? What is the path? Partners in the final round push hard on the second question (right to win), which is where business judgment gets scored.

MECE in 60 Seconds: The Drill That Actually Works

MECE is not memorized, it is drilled. Every day for two weeks, take a random business problem and give yourself 60 seconds to write a tree with two levels and three branches per level. By week two, your tree will cover 80% of the right territory before the interviewer has finished reading the prompt. Below are 6 drill prompts to start with.

• ●"A regional airline's profit fell 20% while passenger volume held flat."
• ●"A European pharma is deciding whether to enter the Indian insulin market."
• ●"Estimate the annual revenue of UPI transaction fees in India."
• ●"A coffee chain's store-level margin is down in tier-2 cities but up in tier-1."
• ●"A B2B SaaS company is losing 30% of customers at month 13."
• ●"A telco operator is deciding whether to bundle streaming with its 5G plan."

5 Worked Profitability Cases

Case 1: European Airline, Profits Down 20%, Volume Flat

If volume is flat and profits are down, the problem is on the margin side, not the demand side. Structure: split cost into fuel, labour, aircraft, and overhead. Fuel spiked 30% in Q3 2025. Labour is under union contract. Overhead unchanged. The diagnosis is fuel exposure, and the recommendation is a staggered hedge on 50% of jet fuel over 18 months, preserving 70% of the margin erosion.

Case 2: Indian D2C Skincare Brand, Profit Down Despite Revenue Up

Revenue grew 40%, profit fell 15%. Structure: gross margin, marketing cost, fulfillment cost. Gross margin held. Marketing cost as a percent of revenue doubled due to Meta ads inflation. The real driver is customer acquisition cost, not product. Recommendation: shift 40% of budget to organic Instagram plus influencer channel, target a 30% reduction in blended CAC over 2 quarters.

Case 3: US Regional Bank, Net Interest Margin Compressed

Structure: net interest margin = yield on loans minus cost of deposits. Both moved in 2024. Loan yield flat because portfolio is fixed-rate mortgages. Deposit cost up because Fed tightening forced higher CD rates. Recommendation: accelerate variable-rate product mix to 40% of new originations within 18 months, and introduce a tiered savings product to anchor cheap deposits.

Case 4: B2B SaaS, Net Revenue Retention Fell from 120% to 98%

NRR equals starting ARR plus expansion minus churn minus contraction, divided by starting ARR. Structure: which of the four components moved? Churn flat. Contraction tripled because mid-market customers downgraded seat counts post-layoffs. Expansion slowed because the usage-based pricing tier has a price ceiling. Recommendation: introduce an enterprise AI add-on priced at 20% uplift, targeting top 100 accounts for a Q2 pilot.

Case 5: Coffee Chain, Margin Down in Tier-2 Cities

Structure: per-store P&L. Revenue per store equals ticket size times transactions. Ticket size flat. Transactions down 15% in tier-2. Labour cost fixed. Rent down 20% in tier-2, not enough to offset. The real driver is declining footfall due to a new competitor (Starbucks expansion). Recommendation: roll out a 30% loyalty discount on the 10am to 4pm window to defend local regulars, reassess in 90 days.

5 Worked Market Sizing Cases

Case 6: Estimate Annual UPI Transaction Fees in India

India processes roughly 15 billion UPI transactions per month in 2026. Annualize: 180 billion. Current fee structure is zero for P2P and 1.1% MDR for P2M on transactions above ₹2,000. Assume 30% of volume is P2M above threshold: 54 billion transactions, average value ₹5,000. Fee pool: 54B times ₹55 = ₹2,970 Cr annually. Sanity check against public NPCI data: within 10%.

Case 7: Estimate Global Google Workspace Licenses

Bottom-up via businesses. ~400M formal businesses globally. Address only knowledge-worker businesses: ~60M. Assume 40% adoption of cloud productivity (vs 60% still on Microsoft): 24M businesses. Average 15 seats per business: 360M licenses. Cross-check top-down against Google's reported 10M paying businesses times 30 seats: 300M. Triangulate to 300-360M.

Case 8: Estimate Number of EV Chargers Needed in Bangalore by 2030

Bangalore has 13M people, ~5M households, assume 30% car-owning: 1.5M cars. Assume EV penetration at 25% by 2030: 375K EVs. Typical charger-to-EV ratio for viable network is 1:10: 37,500 chargers. Split across home (70%), office (20%), public (10%): 3,750 public chargers. Sanity check against current Delhi planning ratios: aligns.

Case 9: Estimate Annual Revenue of McDonald's Breakfast in India

McDonald's India has ~550 outlets. Average 800 transactions per day. Breakfast window is 3 hours out of 14 operating hours, contributing ~15% of transactions. Breakfast ticket size ₹180. Revenue per outlet per day: 800 times 0.15 times ₹180 = ₹21,600. Annual: 550 times ₹21,600 times 365 = ₹434 Cr. Sanity: roughly 7% of India system sales.

Case 10: Estimate Revenue of Paid Newsletter Subscriptions in US

US has 260M adults. Assume 12% are active newsletter subscribers of any kind: 31M. Of those, paid penetration is 8%: 2.5M paid subscribers. Average subscriber pays for 1.3 newsletters at $8/month each. Annual ARPU: $125. Market: 2.5M times $125 = $312M. Triangulate against Substack public data: Substack alone is ~$40M in creator payouts, so market multi-platform total of $300M is plausible.

5 Worked Market Entry Cases

Case 11: Should Swiggy Enter the Grocery Delivery Market?

Four questions. Market: Indian online grocery is ₹80K Cr, growing 25%. Attractive. Right to win: Swiggy has last-mile density, customer base, and brand. But Blinkit has 2-year head start, Zepto has warehouse network. Fit: fits food ecosystem, but requires cold-chain investment. Path: acquire or partner. Recommendation: acquire a tier-2 player at a $300M ceiling rather than build, enter within 6 months.

Case 12: Should Netflix Enter Live Sports Streaming?

Market: global live sports rights are $60B. Growing. Highly attractive. Right to win: global distribution, ML recommendation, but no sports rights expertise and late to NFL, Premier League, IPL cycles. Fit: expands engagement outside original content cycle. Path: bid on regional tier-2 leagues first. Recommendation: target one flagship (NBA League Pass global) and one regional deal in each top 5 market, multi-year.

Case 13: Should a European Pharma Enter the Indian Insulin Market?

Market: ₹6,000 Cr, growing 14%, driven by diabetes surge. Attractive but price-regulated. Right to win: patent expired, so commodity play. Must compete with Biocon, which has cost advantage. Fit: yes for portfolio depth. Path: license to a local manufacturer or build. Recommendation: license-and-distribute for 3 years, reassess greenfield plant decision after capturing 4% share.

Case 14: Should a US EV Maker Enter the Indian Market?

Market: Indian EV car market is 100K units/year, concentrated in sub-₹20 lakh segment. Attractive long term, small today. Right to win: US brand cachet, but Tata and MG dominate sub-₹20L. Fit: requires India-specific battery pack (45-50 kWh). Path: JV with an Indian OEM for assembly, or import CBU at 100% duty. Recommendation: wait 18 months, then JV. The near-term TAM does not justify standalone entry.

Case 15: Should a Bank Launch a Neobank Subsidiary?

Market: 18-35 urban Indian segment, ~180M. Current incumbents: Jupiter, Fi, Niyo. Attractive on digital engagement, low on margin. Right to win: parent bank has balance sheet and lending license, but not cultural agility. Fit: yes for segment capture. Path: ring-fenced subsidiary with independent tech. Recommendation: launch as separate entity with a 3-year profitability runway, separate CEO, and optional exit to merge back after scale.

5 Worked Pricing and Special-Format Cases

Case 16: Price a New SaaS Product for SMB Bookkeeping

Three anchors. Value-based: saves 15 hours/month at ₹500/hour, worth ₹7,500. Cost-based: ₹200 marginal cost plus ₹800 allocation. Competitor-based: QuickBooks India is ₹999/month, Zoho is ₹499. Recommendation: price at ₹699/month with a 3-month free trial, capture the gap between Zoho and QuickBooks.

Case 17: Should a Hotel Chain Introduce Dynamic Pricing?

Structure: who wins, who loses? Chains win ~6% on revenue. Loyal guests lose on peak dates. Execution requires PMS integration and inventory data. Risk: brand perception damage if caught price-gouging on social media. Recommendation: launch dynamic pricing only on 30% of inventory, keep 70% at rack rate for loyalty, roll out over 12 months.

Case 18: Evaluate a $500M Acquisition of a Logistics Startup

Four dimensions: financial (revenue multiple, integration cost), strategic (capability fit, geography gap), operational (team integration, culture), risk (customer concentration, regulatory). Structure matters more than verdict. Recommendation in most McKinsey cases is conditional: yes at $350M, no at $500M, negotiate earnout on 2026 revenue target.

Case 19: Should a Telco Enter Fintech (Payments Plus Lending)?

Jio Payments Bank is the cautionary tale. Market is large but over-banked. Right to win: customer base, distribution, but no balance-sheet risk appetite. Fit: yes for data, no for core. Recommendation: launch co-branded card with partner bank, avoid balance-sheet lending, optimize for fee-based revenue at a 3% blended yield.

Case 20: Design a Go-To-Market for a Rural EV Two-Wheeler

User: rural delivery workers and small shop owners, earning ₹15-25K/month. Product fit: sub-₹1 lakh price, 80km range, locally serviceable. Channel: existing two-wheeler dealers, not new EV-only outlets. Financing: partner with local NBFCs. Recommendation: launch in 3 districts of Tamil Nadu and Maharashtra first, target 5,000 units in year 1 before national rollout.

How Synthesis Separates 3s from 4s

The number one reason strong candidates get dinged in McKinsey final rounds is weak synthesis. A 4 answer is structured as: one-sentence recommendation, two supporting reasons with a number on each, one top risk, one next step. Total time: 60 seconds. A 3 answer rambles through everything discussed. Practice synthesis as a standalone skill, separately from case-solving. After every mock case you do this week, force yourself to deliver a synthesis out loud before you see the next prompt.

Practice a Partner-Led Case in 90 Seconds

Reading these cases is not the same as running one. McKinsey partners will interrupt, push back on your math, and ask follow-up probes you did not prepare for. ZeroPitch gives you a live partner-style case room that simulates the full 25-minute format with adaptive pushback, MECE evaluation, math scoring, and a synthesis rubric. You get a scored report after each round and a stronger-answer example for every moment you lost points.

If you are also preparing for BCG or Bain, start with our consulting interview practice guide, or go deep on case interview math drills before your first round.