2017
I. GETTING STARTED
(start here if you've never done a case interview before)
- Skim (don't read) a case prep book, to orient yourself to the basic structure and likely components of a case interview. I found Victor Cheng's Case Interview Secrets to be honest and practical, but there are tons out there that will do the job.
- Put the book down and practice live as soon as possible. It's critical to become comfortable with the awkward process, develop your own personal appraoch, and most importantly, get practice doing everything quickly and verbally: math, reading a chart, backtracking when you get stuck, synthesizing your recommendations. Mimicking the real interview — quiet setting, 1:1, in-person where possible — is the best way to achieve this, versus reading a book alone or trying to give a mock case to yourself.
- Try to find a case partner to work with consistently. This will keep you both accountable, and there's also a ton to be learned from giving cases to others. Where possible, bother friends who work in consulting to give you case practice and feedback on your approach, since they'll (hopefully) be decent at it and maybe even have some good mock cases to share.
- Once you've done a few cases to get the hang of things, use the tips below to fine tune your technique. Remember that readiness for the interview is defined not by nailing every practice case, but rather by comfort — with your own ability to pick and customize a framework, do some thinking aloud, stay calm when you get stuck, and come up with creative solutions. This comfort will build confidence and free up space (where stress might have otherwise stood) for you to be focused, relaxed, and authentic. Good luck!
II. GETTING GOOD
(a few tips to fine-tune your technique)
- Be friendly and conversational, not overly formal. Your interviewer is looking for someone he/she would not only feel comfortable leaving alone with a client, but also actually enjoy working with (i.e. pass the "airport test" of someone they wouldn’t mind being stranded with in a suburban train station at 11pm on a Thursday). Consulting is an ever-popular career, especially among new grads, and there really is no archetype — so don't pass up your chance to be memorable by stifling your personality.
- Customize your framework to fit the case. The case might throw you a curveball and require you to come up with a new framework to fit the problem, or you can customize a more standard framework by inferring a few things about the industry at hand. Either way, a small adjustment to a cookie-cutter framework is enough for your interviewer to notice and appreciate (on the flip side, it doesn't look so hot if your interviewer catches you trying to force a standard framework when at least one part doesn't apply).
- Create clear, numbered categories in order to keep your thoughts structured. For the initial framework, this might mean forming a few buckets to group your ideas and reading back to the interviewer (e.g. internal: costs, revenues, products and external: competitors, customers, market). Later on in the case, this could be framing your analysis with a sentence upfront: it's always more helpful to lead with "I have two emerging hypotheses about ABC" or "here are three main recommendations to improve XYZ" than to simply launch into the content. This makes it easier for both you and your interviewer to follow your logic throughout.
- Take small cues from your interviewer that hint at the crux of the problem (is it about the customers? the competitors? the geographic segmentation?) and remember that your interviewer is trying to help you solve the case, not lead you astray. After going over your initial framework, pause for a beat to let your interviewer offer thoughts before you begin asking questions — it might be just a simple “great, where do you want to start?” but they may also point out something you missed, which can save you lots of time and frustration later on.
- Make sure every question leads to an actionable answer and communicate clearly why you’re asking it. It’s easy to lose your interviewer or appear disorganized and unstructured if you fire random questions even while staying within an arm of your framework. State explicitly why you’re asking each question and what the answer will allow you to confirm or disprove as you narrow down the problem.
- Work your math out aloud. Practice doing percentages and simple math in your head (e.g. you should know instantly that 40% of 7 million is 2.8 million), round intelligently wherever possible, and remember to double check your answer before confirming. It’s likely your interviewer will help catch small errors if you’ve done a good job of getting your numbers straight and including them in the math thus far.
- Always put your numbers in context by doing quick follow-up calculations and discussing implications after a math segment (e.g. if you just found that you can save 15% by changing suppliers and know that your annual costs are $200M, quickly note that you save $30M in costs each year and compare that to current profit figures to see how your margin might improve). Don’t wait for your interviewer to ask you what the numbers mean.
- Don't overthink the answer. You'll often arrive at a relatively logical set of recommendations at the end of the case, and may be tempted to feel the answer is "too simple" — but the vast majority of cases are NOT rocket science. If you have worked alongisde your interviewer throughout, taken all their cues and guidance, and worked diligently through the math and logic, it's likely you are close to the "right" set of options. The way to actually differentiate your answer, then, leads into my next point...
- Look for next-level insights. Push yourself to think beyond the "simple" answer by looking for insights and domino effects past the obvious numbers, e.g. how might customers and competitors respond to a price change, and how/why does this affect your solution? The interviewer is looking for keen insights that will set you apart from others — a truly stellar response is not just "here's the answer" or "here's the answer and what I think it means" but "here's the answer, 3 things that it implies for the company/financials/customers/supply chain, and a couple things we could do to mitigate risks of those implications."
- Offer rationale and risks to back each of your recommendations. You should always have more than one recommendation or solution to offer, even if one of them is the primary option; try to keep it to 2-3 recommendations overall. It’s okay to give conditional options (e.g. lower prices only if we can be sure manufacturing costs will remain relatively constant), and in fact any recommendation that involves changing prices should be qualified by some thoughtful risks and mitigation. Don’t be afraid to ask for time to jot down your thoughts before giving your recommendation (but try not to take more than 10-20 seconds; definitely shorter than the first pause).
- To find risks, look back at your assumptions. If you’re being asked about potential risks, go back to the initial assumptions you made (or were given) in the original case, unwind them, and see how they might change your recommendations if no longer true.
- Avoid the urge to ramble. This is both a common and easily avoidable pitfall — one where practicing cases live, even giving cases to friends, will help significantly. You can train yourself to pause (and be comfortable with a few seconds of silence), speak slowly, use fewer words to convey the same message, and structure your thoughts before you voice them.
- Bring a high-energy version of yourself to show interest and engagement in the case. Your interviewer has likely been giving the same one or two cases all day, so help them have fun with it — if you're amped about the case, they'll definitely notice and probably get excited too.
- Sounds stale, but RELAX — it really shows. If you find yourself freaking out just before, excuse yourself for a quick two minute power pose in the bathroom (read Amy Cuddy's Presence for proof), take five slow deep breaths, hold your head high, and walk back in — go have some fun with it.