Case Interview Secrets: A Former McKinsey Interviewer
BY VICTOR CHENG
Want to work at a consulting firm like McKinsey & Company? Then you should know that to land a job at McKinsey is 10 times harder than to get accepted to Harvard University.
The good news, though, is that there are smart ways to crack the case interview. Victor Cheng, former McKinsey consultant and case interviewer who aced an incredible 60 out of 61 case-interviews, gives you the interviewer's guide on how to crack the case interview and get job offers not from just one but from multiple consulting firms.
TOP 20 INSIGHTS
1) Recruiters seek candidates who already behave like consultants. Consultants speak for the firm when they engage with a client. Therefore, firms want candidates who pick their words carefully and can factually back every single statement.
2) Besides great analytical and problem-solving skills demonstration, candidates must be able to present their conclusions in a client-friendly manner. This is because clients only accept the recommendations that they can understand. Often, interpersonal skills make the difference between a job offer and a final-round rejection.
3) Interviewers prefer slower candidates who employ a consistent, repeatable process over faster ones who use unrepeatable methods. It is easier to coach for speed, but difficult to improve accuracy.
4) The case interview closely simulates the everyday experience of a consultant’s job. Case interviews involve estimation questions because clients ask them every day. If a candidate doesn't enjoy the interview preparation process, chances are they might not enjoy the job either.
5) Case interviews are hypothesis-driven, not framework-driven. The hypothesis determines the choice of framework. A candidate who masters the critical reasoning behind issue trees performs better than the one who mechanically applies frameworks. The thoughtless application of standard frameworks is a red flag in interviews.
6) If the complete issue tree is shared right at the beginning, it gives the interviewer a sense of how the candidate approaches the problem. If a candidate doesn't do this, the interviewer may conclude that the candidate has not created an issue tree.
7) “Drill Down Analysis” is a systematic method to visit branches and sub-branches in an issue tree. At each branch, the candidate analyzes relevant data to prove or disprove the branch. If a branch is proven false, the candidate revises the hypothesis and creates a new issue tree. The process ends when a hypothesis is validated.
8) In a client interview, a consultant doesn’t spend the majority of the time on the client's problem solution. Instead, the time is spent on the isolation of the causes behind a client's problem. 75% of the time is spent on problem definition. Candidates must never propose a solution until they are confident of having identified and isolated the problem.
9) Candidates often struggle to manage notes. A highly effective way is to keep exclusive sheets for crucial information like hypotheses and issue trees and use separate sheets for calculations. When a branch is eliminated, candidates can show the updated issue tree to the interviewer.
10) Synthesis is a specific way to communicate progress: state an action-oriented conclusion, three supporting points and restate the recommendation. Synthesis is used at the end of the case and when branches are switched in the issue tree. Interviewers highly value this skill in candidates.
11) Smart candidates use their time efficiently and don’t memorize a dozen frameworks. Mastery of just three general frameworks – Profitability, Business Situation and Mergers and Acquisitions – is sufficient to tackle 70% of case interviews. Creation of custom issue trees will take care of the rest.
12) The Business Situation Framework is useful when you need to better understand the qualitative aspect of the context, refine the hypothesis and find efficient ways to test it. The four main branches are Customer, Product, Company and Competition. They can be utilized to analyze cases like a new product introduction, a growth strategy development or a new market entry evaluation.
13) The “Candidate Led Case Interview” is an open-ended format designed to simulate how consultants approach open-ended client problems. This format has been around for decades and every other format is just a variation of this. Therefore, when this format is mastered, the candidate is empowered to excel in all the others.
14) To buy a valuable minute at the beginning of an interview, a candidate can repeat the client’s question back to the interviewer and ask questions to clarify the client's objective. This gives a candidate time to deal with panic and think of an initial hypothesis.
15) Totals and averages always lie. The ability to segment numbers into components is critical for analysis. The best way to understand the implications of a unit is to compare it to a previous time period and the industry norms.
16) In the “Interviewer Led Case” interview, the interviewer chooses which parts of the case are essential and asks specific questions. This can lead to a lot of abrupt switches in the case flow. Excellence in this format depends on the mastery of the issue tree and synthesis. McKinsey has moved almost exclusively to this format.
17) In the “Group Case” interview, candidates are expected to work with competitors to solve a case. Interviewers use this format to identify how candidates collaborate and build on the ideas of others under pressure.
18) In a “Presentation Case” interview, candidates are evaluated on presentations created with given data sheets analysis. The presentation structure follows the synthesis format: groups of slides for the actionable conclusion, three key insights and restatement of the recommendation. Each slide has a data table or chart, labels and a title that conveys the critical insight.
19) To train for a case interview is more like to prep for a bicycle race than to take an exam. This is because it doesn't test what candidates know but rather what they do under pressure. Candidates who get multiple offers have both case interview knowledge and habits to apply it.
20) The only way to get multiple job offers is a disciplined practice. 90% of Cheng’s students who got an offer practiced for 50 to 100 hours. Candidates who clear the top three firms have participated in an average of 50 practice cases. The best way to train is to first watch an expert give a case interview and replicate best practices in mock interviews.
Top-tier strategy management firms like McKinsey and Boston Consulting group rely on the time-tested case interview method to hire candidates. The case interview is a broad term for several methods that test problem-solving abilities. Though, there is a range of formats, the skills required to crack the case interview remain the same.
ACT LIKE A CONSULTANT
The candidates who stand out are those who act as consultants. Interviewers look for independent problem-solvers who don't require supervision. In many cases, clients seek directionally accurate answers, which can be handled through approximate calculations. Being exact will cost additional time without adding value, so candidates are evaluated on their ability to do as little as possible to get the job done. Interviewers prefer slower candidates who can consistently follow a problem-solving process over those who reach the right conclusions faster but with an unrepeatable process. This is because it’s easier to coach for speed than for consistency.
Communication skills often make the difference for those who get offers and those who miss out on the final round. Since clients interpret nervousness as a lack of conviction, nervously presenting the right conclusions will get candidates rejected in the case interview. Every statement made to a client must be confidently made and factually supported as consultants represent the firm. Further, clients only accept factually accurate recommendations they can understand. Therefore, interviewers seek candidates who can present their conclusions in a client-friendly manner.
CORE PROBLEM SOLVING TOOLS
These four tools are used by consultants in their regular work and can be used by candidates in any interview format.
Consultants use the scientific method of hypothesis, experiment and testing for problem-solving. A hypothesis must be stated within five minutes after asking a few initial questions. Not stating a hypothesis can lead to rejection.
2. ISSUE TREE FRAMEWORK
The candidate uses an issue tree to test the hypothesis. An issue tree is a set of logical conditions that, if proven correct, prove the hypothesis. Frameworks are issue tree templates developed by consultants for commonly occurring problems. They must be customized for each case. The hypothesis decides the choice of framework or custom issue tree. Using standard frameworks with no relation to the hypothesis is the single biggest red flag in an interview.
There are three simple tests to check the validity of the issue tree:
The purpose of using an issue tree is to test a hypothesis. Candidates who use a standard framework without stating a hypothesis get rejected.
Issue trees must be mutually exclusive and collectively exhaustive (MECE). Decision-making factors must be categorized into discrete categories that are mutually exclusive. These categories combined must cover all possible options (collectively exhaustive).
A valid issue tree must produce conclusive results. Every important factor must be included in the issue tree and ambiguous factors must be removed.
Candidates must share the full issue tree in the beginning. This gives interviewers a sense of their problem-solving abilities before the analysis. If the interviewer can’t see the entire issue tree, they will conclude that the candidate has not made one.
3. DRILL-DOWN ANALYSIS
This is a process of navigating the issue tree by drilling down branches and sub-branches using data to prove or disprove the branch. If a branch analysis disproves the hypothesis, it’s time to revise the hypothesis and create a new issue tree. This repeated process of drilling down and pulling up ends when a hypothesis is validated. Typically, the candidate refines their hypothesis numerous times in a case interview. Here are some pointers to remember:
Candidates should start with the branch that eliminates most uncertainty by revealing critical information.
A typical case involves roughly 70% quantitative and 30% qualitative analysis. Quantitative analysis is useful to analyze what is causing the problem and by how much. Qualitative analysis helps us understand why decisions were made and how processes work.
The 80/20 rule says that 80% of results come from 20% of data. Candidates must not analyze problems beyond what is minimally necessary and use just the data required. Interviewers consider overanalyzing a situation as highly inefficient.
A simple way to manage case notes is to use two sets of paper. The first set focuses only on the case structure like hypothesis and issue tree diagrams. The second set of papers is a scratchpad for calculations.
Whenever a branch is eliminated, candidates can cross the branch and show the updated issue tree to the interviewer. Involving the interviewer in the problem-solving process makes them more likely to support the conclusion.
The best way to understand the implications of a metric like a customer's variable cost per unit is to compare it to the unit in a previous time period and the rest of the industry.
Synthesis is communicating the results of the analysis in a concise way that is integrated into the overall business context. This is done at the end of the case interview and when switching branches in the issue tree. The standard synthesis template states an action-oriented conclusion, three supporting points and ends by restating the recommendation. Interviewers highly value this format because it presents findings without wasting a minute. Mastery of this skill sets candidates apart from competitors. The best way to master this skill is to record yourself presenting the synthesis, find improvements and iterate.
THREE CORE FRAMEWORKS
Memorizing a dozen frameworks is a bad idea because too much time goes into memorization and the candidate has little proficiency in any framework. In contrast, mastering three general frameworks: Profitability, Business Situation and Mergers and Acquisitions, is enough to handle nearly 70% of cases in case interviews. Custom issue trees can be used to handle the rest.
THE PROFITABILITY FRAMEWORK
This tool is excellent for a quantitative understanding of why a client is losing money. Profits consist of two branches: revenues and costs. While working through the framework, it’s essential to segment and isolate the problem. Profits are segmented to isolate if the issue is revenue-driven or cost-driven.
This process continues to the bottom of the branch, at which point a candidate can ask the interviewer for segmented information like units sold in each region or by demographics. There are multiple ways to segment data, so it’s best to ask the interviewer to share segmented data and allow them to decide the choice of segmentation. Once the issue is isolated, it makes sense to switch to a qualitative framework to understand why it occurs.
THE BUSINESS SITUATION FRAMEWORK
This framework is used for developing a qualitative understanding of the client business, market and industry. It can also be used in the beginning to get basic insights to form a hypothesis. The four components are:
1. CUSTOMER ANALYSIS
Customer Analysis is used to understand customer segments, their needs, how they make a purchase decision and price sensitivity. Different customers prefer different distribution channels. Customer concentration, on the other hand, helps in understanding how many clients exist and how they are distributed. If the customer concentration is higher than the suppliers, then suppliers have market power or vice versa.
2. PRODUCT ANALYSIS
This branch helps in understanding the nature of the product and how, when and why customers purchase it. Further, the analysis focuses on whether it's a commodity or a unique good if there are complementary products or substitutes and the product life cycle.
3. COMPANY ANALYSIS
This branch is useful for a qualitative understanding of a client company.
Capabilities and Expertise: Asking these two key questions can bring out highly valuable insights :
What does this company do well?
What does this company do differently than its competitors?
Distribution Channels: This involves analyzing the company's distribution channel mix and comparing it with competitors' distribution channel mixes and customer preferences.
Cost Structure: The cost portion of the profitability framework is used here to define available strategic options.
Intangibles: A reminder to consider if intangible assets like brand, reputation and culture are relevant to the hypothesis.
Financial Situation: This section is used to analyze both segmented sales and costs.
Organizational Structure: Relevant in cases that involve execution aspects to identify conflicts between structure and strategy.
4. COMPETITION ANALYSIS
This branch usually looks at factors like competitor concentration and market structure, best practices, barriers to entry and regulatory environment.
Competitor Concentration and Structure: Asking how many competitors exist and how big they are gives a sense of how much market power they possess.
Competitor Behaviors: The questions from the product and company sections of the business situation framework can be used to analyze the strategic choice of competitors.
Best Practices: The critical question is, what can the client learn from competitors. If the client can’t beat the competitor on its strengths, an option is to refocus the business on the competitor's weaknesses.
Barriers to Entry: Helps formulate effective market entry plans.
Supplier Concentration: Sometimes, competitors may act in response to moves of a dominant supplier like Intel or Microsoft.
Regulatory Environment: Highly regulated industries might prevent specific actions that make strategic sense for the client. Recent changes in the regulatory environment may create new strategic opportunities.
There is no time to ask every question in every branch. In most cases, candidates get a critical insight into decoding the client’s problem before finishing the entire framework. If this is the case, it's best to pause, synthesize, revise the hypothesis and determine the least amount of information required to test the revised hypothesis.