Case Interviewing

September 19, 2009

A major stressor for a bunch of my students at this time of year is case interviewing. Many of the top recruiters of both UG and MBA students use some form of problem solving interview to gauge candidates’ analytical skills as well as their grace under pressure. I’ll lay out a few of the major themes I review with people as they prepare to go through this process.

What are firms looking for?

  1. Analytical abilities – How clearly can you develop an analytical framework, dissect a problem, create clear alternatives and effectively communicate a recommendation?
  2. Poise – How well do you handle stress and intense questioning? Do you get lost in a sea of data or do you understand what’s important? Do you stand your ground when appropriate or cave under pressure? Do you hard headedly stick to ill conceived positions? How comfortable are you in your own skin?
  3. Fit – Are you one of us? Do you think like we do? Am I comfortable with you in front of my clients on day one?

Each of these is important and falling down on any one is a killer. Keep in mind that you can do really well on all and not get an offer based on how competitive the field is for these roles.

How do I handle the case?

1. Frame the problem

You are typically asked a question or presented and introductory slide or two with data. Make sure you understand the question and then take a few minutes to collect your thoughts and commit them to paper in a logic tree.

Success (failure is the inverse of these):

  • You clearly repeat the question and understand any clarifications the interviewer offers
  • You lay out a reasonable approach


  • DO NOT assume too much. I see candidates try to be too clever and jump to conclusions. This is death in a case. You will waste time, spin wheels and not be demonstrating any of traits 1-3 above.

2. Dig for details

Once you’ve laid out a framework you start asking good questions. This is a narrowing process to eliminate potential problems of avenues that don’t apply. You are trying to show your breadth and depth of knowledge as well as your ability to synthesize in real time.

This is where you are showing your analytical skills. How well do you handle challenges? How well do you respond when what you thought might be the answer has no data and so (obviously) isn’t a useful path?

Also important is how curious and intuitive are you? Stated differently, do you demonstrate an aptitude for being a driving problem solver? Too many people are trying to be formulaic in digging. I call it “wielding frameworks.” Don’t tell me you’ll do a Porter’s 5 Forces. Do it. When I pose you an interesting conundrum, be curious and dig until you’re satisfied you’ve reached the bottom of the information well.

Also, think about what you’re saying here. Many people simply talk too much without good information. You can put waayyy too much out there for your interviewer (or inquisitor) to jump on. Share pertinent information or insights. You definitely want to share your thinking. Just be careful what and when you share it.

More business students than you might think have gotten great grades, but aren’t “savvy”. They can lose sight of what matters in unstructured problems. Solving an undefined case is much harder than getting to the crux of a pricing case when you’re in a pricing class and the case relates to this week’s reading. In that situation, you already have the problem domain narrowed way down. Here you have to figure out the problem


  • Asking good questions that are on point
  • Taking direction and guidance from interviewer. Most are trying to assist you. Don’t treat them as adversaries.
  • Demonstrating curiosity
  • Demonstrating good business acumen/intuition/savviness


  • Not listening to guidance
  • Talking too much in advance of intuition and/or data
  • Failing to follow through on useful paths (this is where “curiosity” matters)

3. Develop the math/reach insight

There is typically a key insight in the case. It might be quantitative – I’ve given you numbers and you reach a financial conclusion (ex: Product line A isn’t profitable) It might be qualitative – based on asking good questions it turns out that you learn a new competitor has entered the market. Either way, you reach an insight that can then drive you to conclusions about sensible recommendations going forward.

It is critically important that you demonstrate comfort with number for those firms that care. Assume McKinsey, Bain, Boston Consulting and Deloitte all care a lot. Do you get what direction the numbers are moving and at what scale? Also – don’t make it harder than it has to be. My emphasis is on understanding. Don’t worry about the 3rd decimal point here. If someone says 190,000, do the math with 200,000. It’s easier and you’ll still get to the right conclusion. Keep it as simple as you can.


  • Obviously, getting the “right” answer matters. But it’s not the only thing.
  • Understanding the insight you’ve reached. Again this sounds obvious, but many candidates don’t understand what they’ve just learned.
  • Demonstrating comfort with the scale and direction of numbers.


  • Thinking you are done. You’re NOT! This is only partial credit. You have to move on to step 4.
  • Blowing the math. People get wrapped up in the math and can fluster.

4. Make recommendations

Take time to think through what to say. It’s an opportunity to regroup and spend 60 seconds jotting down your ideas. You are trying to show the comprehensiveness of your thinking on solutions to the problem as well as the implications of the solutions.

As an example, I use a case that is about profitability. One of the potential recommendations a candidate could make amounts to “eliminate this customer segment because we lose money on them”. Perfectly good recommendation if qualified by some statement like “understanding that I’d want to know more about our fixed vs. variable costs (and then why)”. In essence, do you understand that the financial accounting for profit might be different than the managerial accounting of it? Unqualified you leave the impression you don’t understand the cost accounting implications of what you’ve said and I will proceed to dig until it hurts. On the other hand if you don’t even mention cutting this segment you are missing an obvious potential solution.

The idea is to frame all the available actions that could be promising, walk the interviewer through them and the implications of each.


  • Getting the bulk of the potential recommendations
  • Responding well to challenges of your logic


  • Missing major potential solutions
  • Not understanding the likely implications of your recommendations
  • Thinking your done at “insight”
  • Being shallow

Closing thoughts

  • Practice. A lot.

Successful candidates have typically done 10+ practice cases and I have some students who’ve done 25+. Some people may be clever enough to just stand and deliver w/o this practice. I wouldn’t count on you being one of them. In many case it takes a 3.8 GPA and very high standardized test scores to be considered for a 1st round interview.

A Bain consultant recently told me (paraphrased) that “everyone we interview is super smart. It’s more about how they handle our process and culture. Then it’s how they stack up against other candidates what are all awesome as well.”

Practice can take many forms. Get in the habit of asking your self “why?” a lot. Read the WSJ with a clinical eye and try to understand what’s going on behind a particular story. Find a few friends and grill each other on mini cases etc.

  • Relax and have fun – If you don’t find this at least a little fun, you are probably not well suited to the industry.
  • This discipline is directly applicable to ANY problem solving situation. The same skills apply to work place, classroom and any other environment that requires organizing your thinking to solve a problem.
  • Find contacts at the firm who’ll practice a case with you. They all have slightly different styles. Most alums want to help you. Let them.
  • Resources: The best selling case prep book is Case in Point. It’s a great resource to start with.  Don’t get too wrapped up in the frameworks and drilling on them independently. Practice matters far more past a basic understanding of the method.

Good luck. Let me know what I’ve missed.

Related Post: Case Interviewing Feedback

Performing When the Lights Come On

May 3, 2009

This week we began our cycle of final client presentations and I was reminded of the importance of performing well in the meeting. This is another “obvious point from Phil”, but let me elaborate.

You can have the most dynamic, data driven and compelling story in the world. But if you don’t sell it and respond credibly to questioning in the meeting then you’re dead. Don’t forget that your audience is not just buying into your content. They are deciding whether they buy you.

So what can we do to improve our likelihood of success? Let’s break it down into components:

The Presentation

I won’t spend too much time here as I think this is an entire multi-post series. However, a few important points are worth making.

1)     Make sure it looks professional – If you couldn’t take the time to make it appear decent, why should I take time to listen? 

2)     Be sure to have run it by stakeholders for vetting and input (as appropriate) – You don’t want to be surfacing “new” or controversial information in most cases. You want people to be saying “I agree” and “that will work” etc. Particularly if you are looking for a decision in the meeting you need all “Ts” crossed and “I’s” dotted. Any doubts will send you to “take another look at that and we’ll re-consider this…” hell.

3)     Have organized it logically to tell the story you want to convey based on your audience – Make sure the story flows and builds sensibly. Your audience won’t all be at the same place, so be careful to ensure you’ve given enough context or background. If you are building to one conclusion you organize differently than if you have a series of decisions etc. Never jam up your material with lots of junk slides. Feel free to use your appendix liberally. A general rule of thumb on slides is that if you don’t have 2-3 minutes of discussion per slide (on average) then you should push it to the back.  I’ll write more on this in the future.

4)     Don’t fall in love with your research/data. There is a phenomenon called “the curse of knowledge”. It essentially states when you know something too well you have a hard time summarizing it simply for novices. Never forget you have spent hours, weeks or months thinking about some of your material. Your audience has 30-60 minutes. Bring it up to an understandable level of summary. Also exclude unnecessary charts or data that are “cool” but not relevant to your central story. The appendix can be huge and is great for this content. You certainly want people to understand how much work has been done, but don’t want to distract.

The Delivery

As I mention above, if you deliver your content poorly it will die. You may not get eaten up, more likely you will just be ignored. Your ability to “stand and deliver” will have a big influence on your effectiveness.

1)     Be confident. Lack of confidence is a killer. It makes everyone in the room less sure of what you are telling them and raises unnecessary doubts. If you are not in fact confident, figure out how to seem so. As they say, “fake it ‘til you make it”.  The more you do it, the more comfortable you are.

2)     Understand your goals and be disciplined in what you do/don’t say. You can’t be trying to make 26 points. Pick your 2-3 major storyline elements and hammer them. You should not get to the end and have people say “that was great” and not know what they need to do.

3)     Pace your content appropriately for the level of thought and discussion required. If you have 63 slides and need several contentious decisions made then 60 minutes isn’t enough. Sometimes you are asked to recommend, sometimes to facilitate discussion. These are very different goals and require different structure of content and delivery of material. Plan accordingly.

4)     Be prepared for challenges. It’s important to have thought through who will be in the room (stakeholders) and what each person’s likely interests and objections are. Ideally you’re on top of this enough to have adjusted your slides to address this, but either way you need to be able to respond in real time.

5)     Plan potential responses. For the top likely challenges you can build well formulated responses, even including specific appendix slides. It’s very compelling when you can specifically address these types of challenges. First, you demonstrate that you thought of the issue. Second, you carried the thought through to analysis and built content around it. Third, it potentially allows you to show respect to opposing points of view. The act of building content can convey open-mindedness.

My experience is that if you are well prepared for key lines of questioning then you will receive fewer challenges as the presentation progresses. Basically, they’ve bought that you know your stuff and allow you to proceed. If you can’t address the first several challenges…ouch. It’s going to be a long day.

6)     Understand the room & setting you are in. You need to be prepared for all the little details of staging. Are we around a table, are there 5 or 50 people etc.? There isn’t a universal rule for “what’s best” . But you do need to understand the environment you’ll be in to effectively plan your delivery.

7)     Be respectful in responding. If you lose your temper or are casually dismissive of any audience member you severely limit your effectiveness.

8)     Practice. If the first time you’re going through your material is in the moment then you won’t have anticipated many of the pitfalls inherent in your content. Several dry runs turn up both flaws in logic, as well as slide/content mistakes.

9)     Manage your nerves/Have fun! I personally enjoy the “joust” of presenting and persuading, but I still get nervous. Practice helps this. In addition, I’d encourage you to take the attitude that this is your opportunity to show all your work.

There are many other subtle tips to offer, but if you actually work at the advice above you’ll have less pain and more success. A disproportionate amount of career success comes from how well you deliver in these key situations. You want to be building a positive reputation.

Let me know if you have questions or would like me to dig into any of these areas more.

Structure is Important (Duh!)

February 14, 2009

One of my students just observed (paraphrased) that “sometimes you just need to remember the basics”. The comment came after a class in which we had speakers from McKinsey & Co. present and discuss their approach to structured problem solving.


I have this session annually and it mirrors much of the course content we present in the enterprise, but I still always take something new away from the talk. The simplicity of the basic approach is valuable, but also easy to ignore because it seems so obvious. From a teaching perspective I always need to remember that just because we talked about it awhile ago, doesn’t mean people remember it if you haven’t been re-enforcing a concept or tool.


The high level outline of the method is to 1) define the problem, 2) structure the problem, 3) prioritize issues, 4) conduct analysis, 5) synthesize findings and 6) develop recommendations. Every firm has their version of these steps. I teach similar steps in my class. It’s not rocket science.


Despite this I remain amazed at the extent to which we don’t take all the steps we know we should, finding rationalizations to avoid them “we don’t have time”, “we already know the question” etc.


So how do we avoid the pitfalls of lazy, sloppy or incoherent thinking? Here are a few steps that should help.


First principle: Bring your client & team along for the ride. They have to have a tangible role in each of these steps if you want the highest probability of a useful outcome.


1.      Write the problem or question down. This seems so obvious, but how often do you really commit it to print and get agreement from everyone on what it is.

2.      Determine who the client or audience is and what their interests are.

3.      Work out a clear framework for solving problem or answering the question. I have an earlier post on issue trees you can reference.

4.      Build a plan. Everyone needs to know what they’re working on. Not everything is equally important, so be prioritizing or de-prioritizing as you go based on your judgment.

5.      Then of course, you have to actually do the research.

6.      Develop recommendations that can actually be accepted and used by your client. There are some subtleties in this step.

·        A recommendation your client hasn’t had a part in building reduce the likelihood of success. ”Success” here is defined as they actually do something. Merely liking your work doesn’t meet this standard. The client has to “own” it enough to implement it.

·        Be practical about what is achievable. Don’t tell them about “best practices” they need to implement that they realistically can’t.

·        Don’t just tell them “what”, tell the “how”. A plan with nice ideas, but no implementation insight is mostly useless.


Each can be handled at varying degrees of detail. A six month process improvement project targeting $7 million in savings requires more thought and planning than a one week quick assessment you might summarize the thinking for on a napkin. Use your judgment.


Following good process through the project greatly increases the probability of success. It also reduces stress and increases client satisfaction because they can see where you are.