I don’t know about you, but whenever I finish reading something that could be useful for me to achieve something I really want to achieve, I am always left wondering what to do next.
I don’t want you to get that feeling, so I’ll give you some next steps on how to learn to structure MECE issue trees here. And by learning, I mean really learning it. Learning to build them for realistic, tough case interview problems, not the simplistic ones you find in “case interview prep La-la land”.
The ability to create MECE structures as you need them is an unfair advantage you can have over other candidates. Most people just can’t do it, and they can’t because it’s hard. But you can learn it with the right tools, techniques, and with enough dedication. I’ve learned it and I was VERY POOR at structuring when I started out.
As we’ve seen in previous parts of this article, to craft MECE structures as you need them, you need to master three core techniques and have two additional ones as back-ups. Here’s a handy chart that shows the pros and cons of each of the 5 ways to be MECE.
You can choose to build a structure using one of these techniques, or to put many together in an issue tree format and end up with a beautiful, tailored structure to your problem.
Most candidates will take the easy road. They will memorize a framework or two, or maybe create their own structure that they replicate in every case they get. “It’s what everyone else is doing”, they’ll say. Well, guess what? Everyone else is getting rejected. Less than 3% of candidates join McKinsey, and, ask anyone who’s been through the process, it is no easier to get into BCG or Bain.
I know it’s tempting to take the easy route and try to skip learning this structuring thing. But you know better. You know that the effort required to learn to create custom structures to each case is NOTHING compared to the frustration of going through dozens and dozens of case practice sessions and ending up in the same place. Nothing compared to the grief of getting a series of “thank you for your time, good luck with other firms” at the end of the process.
It is so much better to do your work right now. So much better to actually learn the skills consulting firms are looking for in their recruiting process. To go to the interviews not anxious, but completely sure you’ll be able to solve the case. Not counting on luck, but counting on skill. Not hoping for the best, but knowing you’ll craft your own success and impress your interviewer while showing your personality and truly having a good time.
I’ve been in both situations. I’ve helped other candidates to get from one point to another. It is just so much better.
By reading this article series and understanding the 5 techniques to build customized MECE issue trees for any case you’re already halfway there. But there’s some more work to do to actually perform during the interviews…
You need to put these techniques into practice to master the skill.
I clearly remember one thing when I was first preparing: there was a vast difference between studying for school and for case interviews.
School is largely about memorizing material. You can master material by reading only. But solving case interview problems is an applied skill. What you know isn’t the key driver of performance. Instead, what you do with it is.
In a way, preparing for cases is a lot like learning to play a sport or a musical instrument, or learning to write well. What you do and how you do it are far more important than what “material” you know.
Ask any athlete, musician or writer and you’ll be quick to learn that the key to be a high performer in those fields is to practice A LOT. Not practicing randomly, which is why you need to proper technique, but doing a lot of quality practice. This is what it takes to excel in applied skills, both quantity AND quality.
To learn to structure well, you need to practice. Here’s how you can practice this skill.
First, pick a case.
Any case. Even a case you’ve solved before in the past. Chess masters re-examine games they’ve played in the past and study what they would’ve done differently in hindsight. They did this to become chess masters, not after they did. If you want to be a “Case interview master”, that’s probably a good idea as well.
Second, find as many ways to break down the problem as you can. At least a one from each type of structuring technique. Do it just like I did for the Nespresso’s market share case in part 2 of this series.
In the first cases you practice this you will get fewer possible structures than I got for this case. That’s normal, you will improve with time. But aim for at least five different basic structures.
Third step: pick ONE of these breakdowns to start your issue tree. This will be the first layer of your structure. Since all your structures from your list will be MECE, prioritize according to insightfulness and efficiency. Insight means you’re showing fundamental characteristics of the problem. Efficient means you can prioritize or eliminate whole parts of your structure with just a little bit of data.
There’s no need to overthink step three. There’s usually a few breakdowns that work well for every problem. If you’re in doubt between two, build an issue tree for each and later compare the differences. If you happen to pick the second best, your structure should be good enough.
Step four: build the rest of the issue tree by breaking down each bucket of the first layer. You can use some of the structures from the list of structures you came up with in step two to do that and also create new ones. Make sure they all come from one the five techniques. This guarantees MECEness. You’re gonna get something like this:
Step five is the most important: to evaluate your structure.
This is the hardest one.
To really master structuring you need to create lots of these techniques. But you also need to know you’re doing it with quality. The easiest way to evaluate your issue tree is to have someone do it for you. Ask a consultant or ex-consultant to check it out and point out 2-3 good things and 2-3 bad things about your structure. But this is not always feasible, I get it. It is hard for most candidates to find such a person and harder still to know if they’re being thoughtful in their feedback.
In that case, there are a couple of ways to self-evaluate.
A favorite one is to practice with cases for which you have examples of good answers. If there are explanations of why they’re good answers and what their flaws are, even better. We’ve helped dozens and dozens of candidates get better by using structuring “drills”. These are exercises we’ve recorded answers for. Candidates will get a case question, try to structure on their own and then watch or listen to our answers. By doing this, they can compare what they do against our proposed solution and because we explain our answer thoroughly, they can get the nuances and understand if their structure is good or not (even if they’re different from ours).
A second way to self-evaluate is to compare your structure against the principles of structuring. Imagine you’re evaluating if a driver is skilled or not. You can contrast that against the main principles of driving: does he drive safely? Economically? Is he quick?
The same logic applies to case interviews. If you understand what are the principles of a good structure and why, you can check if it’s good or not. Being pragmatic, you could have a checklist with principles and compare your structure against the checklist.
After evaluating your structure comes step 6: to improve your structure. Improving is quite easy after you’ve evaluated it, and important because this is how you get better.
This may seem like a lot of work to do just for one structure. I know I got anxious to do all six steps when I was preparing. It always felt like time was running and I needed to do more cases.
But it really only takes 20 minutes to do the whole process, and it’s a process that guarantees improvement. You don’t need to be PERFECT in each structure you make and nor should you strive to. You just need to get better.
But shouldn’t I spend those 20 minutes doing more cases?
If your problem is on structuring, doing whole cases won’t help you as much as focusing on structuring.
You will improve as much (or more) by building one custom structure than practicing one full case with another candidate. Practice makes permanent, so either you learn the skill well before you practice or you’ll be making the same old mistakes over and over again. Then you will convince yourself the mistakes aren’t so bad after all. Good luck convincing your interviewer otherwise.
And really, it is so much more time efficient to do these drills! Let’s do some math together.
A quick case takes about 30 minutes. And then you need to give a case to the other person. That’s one hour. If you’re lucky. Decent feedback takes an extra 10 minutes each way – anything less than that isn’t specific enough for improvement. It’s just generic feedback that could’ve been given to any candidate, regardless of performance. You know the kind: “you should’ve been more structured”. Add in the time to find case partners and schedule the meetings and we’re talking about 1.5 to 2 hours spend per case solved.
In 120 minutes you could’ve done 6 structuring drills. Without the risk of getting a case partner that isn’t interested in giving you good feedback or who just isn’t skilled enough to help you. Without the risk of your case partner flaking on you five minutes before. Without the risk of the case you’re given not covering your improvement areas.
If you do only three of these drills every day for a month I plainly GUARANTEE you’re gonna learn to create your own structures from scratch and get your interviewer delighted. That’s 90 structures. Have you ever met a candidate who’s created and thought about 90 custom case structures? I’ve met only a few, and they were all working at MBB.
It takes a bit of time and effort to put these skills into practice. I’ve never told you it was going to be easy. But I did tell you it was going to work. And I know it will because I’ve seen it working with candidates from different backgrounds over and over again. Even those who didn’t have an MBA from a top 10 school. Even those who’d never worked in business before. Even those who had a lot of trouble structuring even the simplest cases.
And now it’s your turn.
I have a challenge for you. Try doing just one of these drills per day for the next five days. Start the first one right now. I’ll even give you the case:
“Suppose Amazon is worried because of a recent surge of theft in its warehousing facilities. How would you help Amazon mitigate this problem and bring it back to normal levels?”
Start with this one. Set up a timer for 20 minutes to build a damn good structure following the six steps of practice. Then if you feel like, practice another one. Or just wait until the next day to do another one. But do at least one of these for five days. If you like it and feel improvement, you can do more per day. And maybe for more than a week.
You’ll see amazing results!
PS: Keep these three things in mind to self-evaluate your issue trees for this and other problems:
- Did I pick the technique that will bring me the most insight to do the first break-down? The first layer of your structure is critical because it determines the rest.
- Am I missing anything? One way to test this is to google the topic of the case (in this case “warehouse theft”), read a couple of articles or news about it, list down the issues/ideas/hypotheses that come up in those articles and see if there’s a place for each in your structure. If there’s not, it’s probably not MECE.
- Try explaining your structure to someone else (a friends, your romantic partner or even your dog will do!) and see if you sound like a human being going through your structure. If you don’t, it’s probably too complex or too “buzzwordy”.