If you are an aspiring data scientist searching for a job, you are going to end up looking in all sorts of places and industries you may not know much about. Recently through a connection, I got the opportunity to interview as a data analyst at a consulting firm. Everything was going smoothly then at the end my connection asked me if I was familiar with a case interview. I played it off, but in reality had no idea what he was talking about. After some research, I learned that it is a business-minded interview specifically administered by consulting firms to test your ability to meet with clients and speak intelligently. There are some specific problems you must practice to be a competitive applicant. I also learned that case interviews for data analysts may only be a little different from a more general case interview. As I prepare for mine this week, it seems like a good topic to blog about, and hopefully I can give you some tips so that at least your mind does not go completely blank like mine did if somebody asks you about it =].
Part 1. Structure
Case interviews are about being asked open-ended questions so that the interviewer can evaluate your thought process. They often leave pertinent information out and expect you to inquire about it or dig deeper and figure it out yourself. Often they may just give you a short description of a company and what they do, then ask you how you would solve some sort of problem. An example could be something like, “Spotify is a streaming music service, that generates revenue by selling ads or selling subscriptions for ad-free listening, they want to expand their market share.” It is then up to you to inquire about the hypothetical company’s costs and revenue and come up with some sort of solution. They are often questions with no right answers; case interviews are about evaluating your process. You can take designated time to think about your answers in the form of a timeout. If you think it's going to take you more than 10 seconds, asking for a pause is advisable. Do not overly rely on this though, you can gain time by reframing the question and asking clarifying questions. Interviews want to see you think about the problem out loud.
In addition to a long open ended question like above, firms may have a variety of other kinds of questions that can vary from interview to interview. Firstly Brain teasers may come up, where they try to trick you with incomplete information or simply want to see if you can make up a creative, intelligent answer. There can be problems based around calculating percentages and estimating market size generally without a calculator. Lastly you may be asked to look at charts or graphs and summarize your takeaways, hopefully as a data scientist you are already pretty decent at this.
Part 2. General tips
- Get partial credit; talk through your thought process.
The Case interview is about seeing how you think about things, if they just hear your answer, they do not get much information about your process. Many of the questions will not have right answers: justify your thinking all the time unless they are speaking or you have taken a time out.
2. Practice doing some percentages, addition, division, and multiplication in your head.
In consulting it's very important to be able to communicate information quickly and intelligently to clients. Here is an excellent resource for some shortcuts that make accounting math a lot easier.
3. Confirm information and ask the right questions
If you are applying for an entry-level analyst position then one of the key things they are accessing is your ability to clarify things with a higher up. After most questions, summaraize your notes outloud and ask if this is in line with their assessment of the situation. You can also employ some flattery here (as you should in any interview) and tell them what an interesting case this is at this point. In addition to making you look like a better candidate, it will give you time to think through things and come up with some questions.
In my research, I have found that often the difference between a good candidate and a great candidate is the quality of their questions. To be a good candidate you have to ask plenty of questions but to be a standout you have to ask the right questions. Always pull at threads and diagnose things. This video is an in depth look at what separates good, okay and bad questions.
4. Review your business vocabulary
If you are not a business student like me, this step might involve a lot more learning than reviewing. Knowing terms like revenue, costs and the basics of business structure is invaluable.
Part 3. Data analyst specifics
There is a lot less info on the web about the kind of questions analysts should expect, but generally, it seems like there are more likely to be charts and graphs during the interview and more of a focus on data analysis. In many ways the prep is very similar, just expect questions focused on your industry. If the consulting firm is focused on organizing big data, don’t be surprised if they ask you how to tailor revenue streams based on data analysis or what you can even offer a client by reorganizing there data.
One thing I found with a lot of the charts and data questions I looked at was that they often present competing hypotheses to some extent. For example in this question about the education policy of a fictional eastern European country, they prevent evidence that increased spending improves education, but also that it does not. Be very careful about your declarative statements. Do not be over decisive based on glancing at the data. Unless you can run a difference of means test in your head, they are looking for interesting lines of thoughts, not definite evidence.
Part 4. Amazing Resources
Case interviews have a well-established formula, as such, there are a ton of great resources out there. Here are links to the ones I found most helpful.
Case Interview 101 - Basics, Examples, Prep Guide | MConsultingPrep
"Case Interview" is the cornerstone of consulting recruitment, playing a decisive role in final results. In 30 minutes…
How to Prepare for Business Case Interview as an Analyst?
As a data analyst or data scientist, we not only need to know probabilities and statistics, Machine learning Algorithm…
Interview prep, industry insights and career guidance for consulting and product management. Subscribe to get weekly…