YT YouTube2Text

YouTube2Text → Transcripts → Case Interview Fundamentals: Lesson 1 [The 6 Building Blocks of Case Interviews]

Case Interview Fundamentals: Lesson 1 [The 6 Building Blocks of Case Interviews] — Transcript

by CraftingCases · 10,364 words · 1,605 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00Hey there. I hope you're enjoying Case
0:02Interview Fundamentals. This is Bruno
0:04here. And in this video, I'm going to
0:06show you a system so that you can answer
0:09any type of case interview question,
0:11regardless of case type or which firm
0:14you're interviewing with or any other
0:17variable. We call it the six building
0:20blocks and it is the foundation of the
0:22Crafting Cases approach to case
0:25interviews. If you learn this well, it
0:27will allow you to answer any type of
0:30case interview question, even the ones
0:32you haven't seen before. It will allow
0:34you to show your interviewer that you
0:37can think like a true consultant without
0:40relying on memorizing scripts and
0:42frameworks like other people do. And it
0:45will allow you to do the bulk of your
0:48preparation by yourself on your own time
0:51because sometimes you just don't have
0:53access to good people to practice with.
0:56And you're going to learn how to do all
0:58that by the end of this video.
1:01But first, a warning. This video is
1:04pretty long. And even though it's long,
1:08I hate long videos, too, but it's really
1:11really important. It's filled with
1:13concepts and I'm going to just give you
1:15an outline of the video with the times
1:18of when things happen just so you can
1:21have an idea and so you can jump
1:23straight to whatever parts uh you want
1:27to go. But I do recommend that you watch
1:30this whole video. In the first third of
1:32this video, you're going to learn why to
1:35be outstanding, you need to do different
1:37things than to get merely good at case
1:40interviews. And you want to be
1:42outstanding if you want to get the
1:43offer.
1:44You're going to learn why Case in Point
1:46or Victor Cheng or methods like these
1:49ones will let you down
1:51in your journey to be outstanding. And
1:54you're going to learn about the six
1:55building blocks as an alternative
1:57approach. Then in the second third of
2:00this video, you're going to learn what
2:02is each building block with examples of
2:05questions that uh are related to each
2:08building block. And finally, on the last
2:11third of the video,
2:12I'm going to show you two examples of
2:15how to use the building blocks to solve
2:17real cases, how to tie them together
2:20into
2:21real case solutions. All right, so
2:24because this video is long, go get
2:26yourself a cup of coffee. And before we
2:30do any of that, let me start with an
2:33analogy.
2:34And it is an analogy about the building
2:36blocks of football.
2:39For the Americans listening to this, I'm
2:41going to be talking about the European
2:42kind of football, which you call soccer,
2:45but you could think of a similar analogy
2:48for American football or for any other
2:51sport like volleyball or basketball or
2:54golf or actually for any skill. So
2:57guitar playing or poker or whatever. And
3:01the analogy would work well, too.
3:05So if you want to get really good at
3:07football, you can't just play it. You
3:10need to learn the building blocks. So
3:13when you're a beginner, you might just
3:15want to learn the basic rules of the
3:17sport and then play a few matches of the
3:20game.
3:21In the beginning, this is not just what
3:23you'll feel like doing the most.
3:25It is also the best way to improve,
3:28probably. You know, you want to learn
3:29the basic rules, play a few matches just
3:32to see how playing the game feels like.
3:35No one wants to practice the
3:37fundamentals of a sport that they barely
3:39know if they like to play or not. You
3:42know, just give me the ball and let's
3:44have some fun playing it, goddammit.
3:46But after a while playing, you'll see if
3:50you like the game or not, if you want to
3:52get good at it or not. And not only
3:55that, you'll also notice that you don't
3:58really play that well. This happens to
4:00all the beginners. And at this stage, if
4:04you're serious about learning to play
4:06well, if you like the game, you want to
4:08keep playing it and you want to play
4:10well,
4:11you'll have to give some attention to
4:14learning the fundamentals. How to
4:16receive the ball, how to pass the ball
4:19well, how to shoot the ball, how to
4:21cross the ball, how to dribble, how to
4:24tackle, how to position yourself in the
4:27field, and so on.
4:30You can call these fundamentals the
4:32building blocks of football. They're the
4:35basic individual skills that every
4:37player performs.
4:39Any player can practice these skills in
4:41isolation. And what makes a player
4:44better than another is that they can
4:46perform one or more of these skills
4:48better than the other. Either that or
4:51maybe they can tie these skills together
4:54with more skill. So what makes one
4:55football player better than the other?
4:58Probably the fact that they can pass the
5:00ball better or they can shoot the ball
5:02better or they can uh dribble and tackle
5:05better or or they can position
5:07themselves in the field better.
5:09And while you can go from bad to okay in
5:14football just by learning the rules and
5:17playing more and more matches,
5:19you can't get to good, much less to
5:22outstanding, if you're not very good at
5:25dribbling and at passing the ball and at
5:27shooting the ball and in the other
5:29fundamentals. And the same is true for
5:32case interviews.
5:34In this video, you're going to learn
5:36what are the six building blocks of case
5:38interviews. And throughout the course,
5:41you're going to learn how to perform
5:43well in each one of them. We've
5:45dedicated one whole module for each
5:47building block so that by the end of
5:49this course, you're going to be able to
5:51solve each type of problem like a real
5:53McKinsey, Bain, or BCG consultant.
5:58These building blocks are the
5:59fundamentals of how to solve a problem
6:01like a consultant. So they're not just
6:03the building blocks of case interviews,
6:06they're also the building blocks of the
6:08consultant profession, at least when it
6:11comes to problem-solving approaches. So
6:14mastering these building blocks is
6:15necessary if you want to go from okay to
6:18good and from good to outstanding. And
6:21believe me, if you want to get the
6:24offer, if you want to get the job offer
6:26at these firms, you want to get to
6:28outstanding or at least to very good.
6:31But before we move on,
6:34I want you to understand how most people
6:36learn cases.
6:37What are the two most used alternative
6:41approaches? And I want you to see what
6:44these popular approaches are strong at
6:47and where do they fail. And the reason
6:49why I want you to see that is so that
6:51you can judge for yourself if our
6:53approach is the best one for you or not.
6:57So let's take a close look at these two
7:00popular approaches, which is what 95 or
7:0398% of people out there are using and
7:06you're probably also using one of these.
7:09Approach number one is learning by type
7:12of case. So you might want to learn
7:15profitability cases and then move on to
7:17M&A cases and learn those and then
7:20market entry case and learn those. And
7:24the most popular resource that teaches
7:26that way is Case in Point.
7:29It is the approach that most people
7:32start with.
7:33And you know, it feels good to just read
7:35Case in Point and get a sense that you
7:38just need to know what type of case your
7:40interviewer is giving you and then
7:42matching it with the appropriate
7:45framework. And this brings us to the
7:48biggest advantage of this approach,
7:49which is it's really easy to start the
7:52case. You just match the case type with
7:56whatever way you use to start that case
7:59type. But it also limits you on how well
8:03you start the case. So let's talk about
8:06the drawbacks of this approach. The
8:08first one is you can't learn all types
8:10of cases.
8:12So what I see in candidates who use this
8:14type of approach is that they start with
8:16estimation, profitability, M&A, then
8:19they move on to market entry and growth
8:22cases and then pricing cases and then
8:25they start talking about product launch
8:28and then they realize that there's 15
8:31different types of uh public sector
8:33cases just in last year's Wharton case
8:37book and then they also have heard of
8:40this guy who got a case on uh
8:43debureaucratization
8:45and deregulation and then they get crazy
8:48because there's an infinite number of
8:50types of cases. And if you're learning
8:52by type of case, you're you have to
8:54learn about an infinite number of cases
8:57if you want to make sure you get all
8:59cases right. Now, the second drawback of
9:02learning by type of case is you can't
9:05learn everything that could happen in
9:08each type of case. So when you learn
9:11this way, what you're basically doing is
9:14you're basically treating every case of
9:16the same type as the same thing. So
9:19you're treating every market entry case
9:21as the same thing. You're treating every
9:24profitability case as the same as all
9:27other profitability cases.
9:29And you're also not taking care of what
9:33to do after you structure and start the
9:36case. So
9:39when you learn by case type, you're
9:40basically saying, "Well, this type of
9:42case is always the same way." And this
9:45is not true. Interviewers will actively
9:47look for ways to make it different. And
9:51the third drawback of this
9:54approach is that a lot of types of a lot
9:56of cases are mixed type or it's not of a
10:00type that's obvious up front. So, you
10:03might get a case that a company is
10:05deciding or not whether they should
10:08acquire another company because this
10:10other company is launching a new product
10:12in a market that the first company is
10:15really interested in entering. So, is
10:17this an M&A case or a market entry case
10:21or a product launch case? It's really
10:24hard to know.
10:25And then there's cases that you can't
10:27even know the type beforehand because
10:30the problem is not well defined it. So,
10:33this approach of learning by type of
10:36case is helpful in the beginning because
10:39it's really easy to start the case,
10:41which is where most beginners have
10:43trouble with,
10:44but it doesn't help you get better and
10:47it doesn't help you in real interviews.
10:50And most candidates abandon the Case in
10:52Point methods at some point in their
10:54preparation, usually way before their
10:57interviews. And they do so for good
10:59reason. It doesn't really work that
11:01well. Just ask any seasoned candidate
11:04who has gone through the process before
11:06and they're going to tell you that, you
11:08know, it's helpful in the beginning,
11:09maybe, but
11:11it doesn't help you get better and it
11:13doesn't help you in real interviews. And
11:17where most of these people find solace
11:19later is with approach number two,
11:22learning by stage in the case. So, they
11:26might want to learn how to open a case,
11:28how to analyze the case, and then how to
11:31close the case. And this is best
11:34represented by Victor Cheng. So, while
11:37approach number one makes the beginning
11:40of the case variable, so you have
11:42different frameworks
11:44for different types of cases and they
11:46change in the beginning of the case uh
11:49and then they leave you wondering what
11:50to do next,
11:52with Victor Cheng, he inverts it and he
11:55fixes the beginning of the case, so he
11:57just has two main frameworks, the
12:00business situation and the profitability
12:02framework, and he tells you to focus on
12:06what stage of the case you're on and
12:09what's to do next. And here comes the
12:11major advantage of this approach. It's
12:14easy to understand the mechanics of how
12:17a typical case unfolds. So, it's easy to
12:21understand what to do, what to say to
12:23open the case, and then how to open it,
12:26and then once you present the framework,
12:28how to move on to the more analytical
12:30part of the case, and once you've done a
12:32bunch of analysis, how to synthesize
12:35those and present a recommendation. So,
12:38it's really easy to go through the
12:39mechanics of it.
12:41But, this approach also has several
12:44drawbacks. The first one is that a case
12:47interview is not a chess game. It
12:49doesn't always start and end the same
12:52way. And the way you analyze the
12:54different cases, so the middle is
12:57completely different. Some cases have a
12:59lot of numbers. Some cases have no
13:00numbers. Some cases have a lot of
13:02charts. Some cases have no charts. Some
13:05cases you should use hypothesis a lot
13:08and others you shouldn't. And a lot of
13:11cases, a lot of real cases, don't start
13:14in the beginning and they don't end in
13:16the end. So, interviewers will get a
13:18chunk of a problem-solving session that
13:21they did last week or a year ago and
13:24they'll just present that to you and the
13:25case will have no conclusion or you just
13:28start from the middle of the case and
13:30start from from the charts, for example.
13:33A second drawback of this is that
13:36within each stage of the case, there are
13:39different skills and tools to use. And
13:43while Victor Cheng mentions these tools
13:46and techniques, so custom issue trees,
13:48how to generate and test hypothesis, and
13:51other skills like that, he doesn't show
13:54you how and when to use each of these
13:57skills. So, he focuses too much on the
14:01mechanics and on the process and he
14:03doesn't show you how to do it, how to
14:05build your own frameworks, how to create
14:07an issue tree, how to prioritize.
14:10And the third drawback is that it makes
14:13your thought and your communication too
14:16mechanical or as many candidates say,
14:19too robotic.
14:21So, he fixes
14:23the beginning of the case into basically
14:25two frameworks and then he says that
14:27every single case or almost all cases
14:31are to be solved in these three steps.
14:34You open, you analyze, and you close it.
14:36So,
14:37you don't have to be a genius to realize
14:39why so many interviewers complain that
14:42every single candidate sounds the same
14:45nowadays.
14:46And because most other methods that you
14:50see out there, that are not Case in
14:51Point and are not Victor Cheng's, are
14:54copies of Case in Point and Victor
14:57Cheng, there's not much of a way out of
15:00this. You're either learning by type of
15:02case or you're learning by stage of the
15:05case.
15:06And these approaches are helpful in the
15:09beginning, but they fail to help you as
15:13you get more advanced.
15:16And you might be wondering, why is that?
15:18That's something I wondered for a long,
15:20long time. And I found the reason. The
15:23reason is they skip the fundamentals.
15:26Right? You wouldn't help someone get
15:29good at sports by lecturing them about
15:33all the different types of sports. And
15:35you wouldn't help someone get good at
15:37football, for example, by showing them
15:40how there are three different parts in a
15:4290-minute match. There's the beginning
15:45of the match, there's the middle of the
15:47match, and there's the end of the match.
15:50The way to
15:52get better and get more advanced and to
15:56get go from okay to good and from good
16:00to outstanding is to learn the
16:02fundamentals, the building blocks. This
16:05is true for sports and this is true for
16:08case interviews.
16:10So, let's take a look at approach number
16:12three. Let's take a look at our
16:14approach. Let's take a look at the six
16:17building blocks of case interviews,
16:19which are the fundamental
16:20problem-solving techniques that real
16:23consultants use in their jobs. There are
16:26three building blocks related to
16:29structuring the problem.
16:31These are brainstorming,
16:34frameworks, and hypothesis testing. And
16:38these will help you structure the
16:39problem. They will help you ask the
16:41right questions. And there are three
16:44building blocks related to analytics.
16:47These are estimations,
16:49quantitative analysis, and chart
16:52interpretation.
16:54These will help you analyze the problem,
16:56answer those questions that you asked
17:00when structuring the case. So, whenever
17:03the interviewer asks you a question in a
17:06case interview, that question is going
17:09to be related to one of these building
17:11blocks. Whenever you do a task during a
17:14case interview, it is going to be
17:17related to one of these building blocks.
17:20This was tested over thousands of
17:22interviews and no single question, no
17:25single task
17:27uh being performed in any case interview
17:30from any firm that we've seen in the
17:32past
17:33uh 3 years in thousands and thousands of
17:36interviews was not related to one of
17:38these six building blocks. So, what this
17:41means is if you master the six building
17:44blocks, you're going to be able to solve
17:46any case interview as a candidate and
17:49because this comes from the consulting
17:51job, you're going to be able to solve
17:53any of your clients' problems as a
17:56consultant.
17:57Each building block can and should be
18:00solved using a best practice, a
18:02step-by-step
18:04method, which you learn throughout this
18:07course. You're going to learn a
18:08step-by-step method that bulletproofs
18:11your answer for each of the six building
18:14blocks through this course.
18:17Using these systematic processes is how
18:20you assure that you're showing your
18:22interviewer that you can think like a
18:23consultant. And because
18:26all of this this this building blocks
18:29and the step-by-step methods, because
18:31all of this comes from the consulting
18:33job itself,
18:35they are a reliable way to tackle case
18:38interviews. So, if you use these
18:40methods, you get several advantages over
18:43other methods. The first one is that
18:46this is thorough and works for any case
18:50interview. So, every case that you face
18:54will have these elements in them. It It
18:57doesn't matter if it's a profitability
18:59case, if it's an M&A case, if it's a
19:01case at McKinsey, if it's a case at
19:04Bain, at BCG, or at your local boutique
19:08firm. Because this comes from the
19:09consulting job, you can solve any case
19:12using it.
19:14Each part of
19:16uh the case or each of these building
19:18blocks can be learned and practiced in
19:21isolation. So, suppose you're learning
19:24football. If you're not really good at
19:26shooting the football, you can practice
19:28that in isolation. You don't have to
19:30gather 21 other players to practice your
19:33shooting. You can just get the ball and
19:36practice shooting. If you have a a
19:38goalkeeper, that's better, but you don't
19:40even need a goalkeeper to do that. Same
19:43here. If you're not really good at
19:45hypothesis testing or in estimations or
19:48in any of the building blocks, you can
19:50practice that in isolation.
19:54Third advantage, it does not
19:56oversimplify cases, but it makes them
19:59simpler to solve. So, other approaches
20:02will oversimplify
20:04the cases and will help you in some
20:07aspects, for example, how to start a
20:09certain type of case or how to go
20:12through the mechanics of solving a case,
20:14but will harm you in other aspects. And
20:17they will harm you by having you believe
20:19that you can just memorize frameworks
20:22for a bunch of cases or you can just use
20:26two types of structures and then repeat
20:28a pattern in every case you do. And
20:31these things are not true and that's why
20:32these other methods might harm you.
20:35Our methods does not oversimplify the
20:38cases, so it will never harm you, but it
20:41does make everything simpler. Fourth,
20:44there's no memorization needed. It is a
20:46system to help you think. So, we won't
20:50teach you some tips and tricks for each
20:52building block or give you a bunch of
20:54scripts and frameworks to memorize.
20:57Instead, we're going to show you how to
20:59think through each type of problem. Each
21:02building block is a different type of
21:04problem and we're going to teach you how
21:06to think through each of them.
21:09Fifth, it comes straight from the
21:11consulting job, so it doesn't sound
21:13robotic. Real consultants also build
21:16frameworks. They also brainstorm
21:18hypotheses. They also do Excel models to
21:22do quantitative analysis. They also
21:24interpret charts. These techniques come
21:27straight from the consulting job. We
21:29didn't make up the step-by-step methods
21:31for each one of them. We got them from
21:33the consulting job, from what we and
21:36other McKinsey, Bain, BCG consultants do
21:39in their daily jobs.
21:42And your interviewer will use these
21:45techniques as well, consciously or
21:47probably unconsciously. So,
21:50when they see you uh solving your case,
21:54they're going to see themselves solving
21:56their clients' problems and this makes
21:58your solution much, much better. You
22:01won't sound robotic. You'll sound like a
22:03real consultant. Now, this has one
22:06drawback compared to other methods.
22:10And the drawback is there are no
22:13shortcuts. You really need to understand
22:16what you're doing in each building
22:18block. And it does take work to learn.
22:21We've distilled the the essentials in
22:24the 7-day course, but I'd be lying to
22:27you if I told you that you can master
22:30uh and become a great consultant in just
22:33a week.
22:34It takes understanding and it takes
22:38thinking things from first principles
22:40and it takes practice and in this
22:43course, I'm going to lay or Julia and I
22:46are going to lay the foundation of each
22:48of the six building blocks, but you got
22:51to understand and practice and put into
22:54action. It's not a shortcut, but it's a
22:57really solid, strong method
23:00that will help you think like a real
23:03consultant and show your interviewer
23:06that you can do that. Now, I'm going to
23:08show you what each of these building
23:09blocks are with a few examples of case
23:12interview questions that you might get
23:14in each of them. Then, after going
23:16through all the six building blocks, I'm
23:18going to show you how to fit them all
23:19together to solve any case interview.
23:22Candidate-led, interviewer-led, it
23:25doesn't matter. In a couple of full case
23:27examples. So, let's start with the
23:30structuring related building blocks
23:32first and the first of them is
23:34brainstorming. What is brainstorming?
23:38Brainstorming is using issue trees to
23:40build a structured list of hypotheses
23:44and ideas. It is not throwing out a
23:47random list of ideas or hypotheses
23:50without structure and priority. So, in
23:54every interview, you're going to need to
23:56work with hypothesis of why something's
23:58happening and ideas to solve the
24:00problems that appear. And you'll have to
24:02come up with them just like consultants
24:05have to come up with them in real
24:07projects. And the way consultants come
24:10up with them in real projects is in a
24:12structured way. So, it's not getting
24:16around a whiteboard and just throwing
24:18ideas unsorted. It's in a structured
24:21way. And this isn't like what most
24:23candidates do. Whenever interviewer ask
24:26candidates
24:27uh what are some hypotheses that you
24:29have for for this phenomenon that's
24:31happening or what are some ideas that
24:34you have to solve this problem, what
24:36most people do is to just list a random
24:39list of the ideas that come up in their
24:41heads or of the hypotheses that are off
24:44the top of their heads. And
24:46this is a mistake if you want to get a
24:49job in consulting. In the brainstorming
24:51module, we'll teach you what to do
24:53instead.
24:55So, let me just show you some examples.
24:57Uh and all of these questions, these
25:00examples, they could be the beginning of
25:02a case or they could appear in the
25:04middle of a case.
25:06So, an iron ore mining company has seen
25:09its revenue drop 30% in the last year
25:12despite heavy investments. What are
25:15plausible hypotheses that could explain
25:18that? So, this is something in the
25:20beginning or in the middle of a case, it
25:21doesn't matter.
25:22Uh they're asking for a list of
25:24hypotheses and that's why it's a
25:26brainstorming question.
25:28A large university wants more applicants
25:31for their undergrad program next year.
25:35What are some ideas of what they could
25:37do to get that? Again, a list of ideas,
25:39brainstorming question.
25:41Apple wants to launch a cheaper version
25:44of iPhone. What are the risks? Here,
25:47it's a list of risks, but you could call
25:49the risks an idea or a hypothesis of
25:52risk. And
25:54all these questions could be in the
25:55beginning of the case or in the middle
25:57of a case and
25:59most cases have between two and four
26:02brainstorming questions. So, they're
26:04very, very frequent. And interviewers
26:08tend to ask them spontaneously, so most
26:10candidates are not even aware that this
26:12type of question exists. They don't
26:14practice for them in mock interviews and
26:17because of that, they come underprepared
26:20and they suffer the consequences. So,
26:22really important building block, really
26:24important if you want to learn how to
26:26structure hypothesis, ideas and cases in
26:30general. The second building block
26:32related to structure are frameworks. So,
26:35you're you're probably well aware what
26:37frameworks are in terms of, yeah, I know
26:41some frameworks, I know the pricing
26:43framework from book X or whatever, but
26:46you might not know that frameworks are
26:48decision-making plans. They're a list of
26:52the questions that you need answered to
26:55make a decision and a systematic way to
26:57make sense of this list of questions.
27:01And the skill being tested here by
27:03interviewers is how you create this
27:06plan, not if you know a pre-made plan
27:09from a case interview book. So,
27:11frameworks are usually the initial
27:13question of the case, but don't count on
27:16that. It is common that interviewers
27:18start with a brainstorming question
27:20about the topic of the case just to warm
27:23you up and then go on to the decision to
27:26be made. And when you have a decision to
27:29be made is uh when you know that you
27:32should make a framework. And I've also
27:35seen framework questions at the end of
27:37the case, after you've solved the
27:39client's initial problem. You do the
27:41case, you structure the case, you solve
27:43the case, you give the client a
27:45recommendation and then the interviewer
27:48will pose a second, more difficult
27:50decision that follows up the first
27:52decision just to see how you structure
27:54it. And this is similar to how real
27:57projects work. Each answer to the client
28:00reveals new questions to be answered.
28:04So, some examples of questions here. How
28:06can Nestle create a long-term growth
28:09strategy? So, this is a decision that
28:12they need to make. How should we go grow
28:14long-term?
28:16Help a health minister combat childhood
28:19obesity. So, again, this health minister
28:21needs to create a plan.
28:23He needs to make a decision how to
28:25combat uh the this ill and he needs a
28:29plan for that. How would you help a soft
28:31drinks company increase its profits?
28:36So, again, they need to make a decision
28:38on how to increase profits.
28:41So, in the frameworks module, you're
28:44going to learn how to create your own
28:46frameworks, so your own decision-making
28:49plans for any type of case, regardless
28:52of you've seen the type of case before
28:54or not. Uh in this methodology, in our
28:57approach, you don't need to have seen
28:59the case before in order to structure it
29:02well. And you're going to learn how to
29:04create these frameworks, these plans to
29:07make decisions for any type of decision,
29:10any type of situation.
29:13And finally, hypothesis testing, the
29:16third structuring building block.
29:18So, when you create your frameworks,
29:21when you brainstorm your issues, you're
29:23going to generate a lot of of hypotheses
29:27for any case you solve.
29:29But what good is it to have these
29:32hypotheses if you can't test them?
29:35Hypothesis testing is at the core of the
29:38consulting way to solve problems. So,
29:41this building block serves to test the
29:44hypotheses that you've organized and
29:47you've prioritized when brainstorming or
29:50when building your frameworks. Every
29:52brainstorming is an organized list of
29:54hypotheses and every framework is a list
29:57of questions that you need to answer.
29:59And these questions derive from certain
30:02hypotheses that you have in your head.
30:05And the hypothesis testing building
30:07block serves to test this hypothesis.
30:10And these tests need to be rigorous but
30:13also practical. So, that's the hard part
30:16of hypothesis testing. You need to be
30:18precise and rigorous
30:20and very scientific
30:22in the way you test the hypothesis but
30:24it also needs to be practical and you
30:27have to be able to test them as quickly
30:29as possible. So, some examples of
30:31questions that could come up in an
30:33interview.
30:35Is passing a law to increase physical
30:38activity in schools a good idea to
30:41reduce childhood obesity? How would you
30:44test that? So, you have a hypothesis of
30:47that passing this law is beneficial to
30:49that goal and the interviewer is asking
30:51you how would you test it?
30:54Would it be worth it for Nestle to start
30:57a super premium line of baby food
30:59products in order to grow sales in
31:02Europe? So, you could approach this
31:05question as a framework. It's kind of a
31:07decision that they need to make but the
31:10question is would it be worth it?
31:11Probably financially. So, how would you
31:13test the hypothesis that they will make
31:16money by doing this?
31:20Will
31:21will the building of nicer dorms
31:23increase the number of applicants of a
31:26university's undergrad program? So,
31:28again, a hypothesis that has a
31:32a potential cause, building nicer dorms,
31:35a potential effect, increasing the
31:37number of applicants. And they don't
31:40know if the cause will cause the effect
31:42and they need to test this hypothesis
31:45and they need you to come up with a test
31:47for that.
31:48So, a final word related to structuring.
31:52People often notice that there is some
31:54overlap between brainstorming and
31:56frameworks, between frameworks and
31:58hypothesis testing. And while there are
32:01technical differences
32:03and you can go into the definition of
32:05each one to figure it out, you can
32:08indeed solve many of the problems that
32:11interviewers will give to you using more
32:14than one of these techniques depending
32:16on how the question is framed. So, for
32:18example, you could tackle example number
32:20two of this list with a hypothesis
32:23testing technique or with a framework
32:26technique.
32:27And the reason for that, the reason why
32:30you can do that is that the the three
32:32things, brainstorming, frameworks and
32:34hypothesis testing, are structuring
32:37techniques. So, just to give you an
32:39example, in football for example,
32:41there's often a blurred line between
32:44whether a ball's being passed or crossed
32:47or whether a ball's being crossed or or
32:50or was shot to the goal. But the point
32:53is can you get the ball from point A to
32:56point B? There's no need to have a
32:59lengthy discussion and conversation
33:01about whether the ball was technically
33:04passed or crossed or shot. It doesn't
33:07really matter. Was the ball
33:09in point A? Did it go to point B? And
33:12the same thing happens here. Your
33:13interviewer doesn't care whether it's a
33:16brainstorming or a framework question.
33:19If he didn't go through this course, he
33:21probably doesn't even know these names
33:23formally or the definitions that I'm
33:26showing you here. They do it
33:28unconsciously.
33:30So, in the end, they just want to see if
33:33you can structure the problem.
33:36And the way you can guarantee that you
33:39will be able to structure any problem in
33:42any situation is that you master these
33:45three techniques, brainstorming,
33:47creating frameworks and hypothesis
33:50testing. All right?
33:52Now, while structuring the problem well
33:54is super important because it helps to
33:57ask the right questions in the right
33:59order, when you start testing your
34:02hypothesis using the hypothesis testing
34:04building block, you need some questions
34:07answered. And one of the main tools
34:09consultants use to answer these
34:12questions is to analyze numerical data.
34:15And there are three building blocks for
34:17that as well.
34:19The first one is estimations.
34:22Estimations are used to quantify
34:25something when you have no data or when
34:27you have insufficient data by using
34:30logic and some assumptions.
34:33And it all it is also used to see if you
34:36have what it takes to model scenarios in
34:40consulting. So, you can have estimation
34:43or market sizing cases on their own or
34:46within larger business cases or public
34:49sector cases. And can they can be very
34:52easy and simple or they can be very
34:54complex and difficult. Now,
34:58let me dig it a bit deeper into the
35:00modeling of scenarios. Consultants
35:02typically model the future growth of an
35:05industry or how much they're going to
35:08save or their clients are going to save
35:10due to productivity increases caused by
35:13certain changes using spreadsheets and
35:17spreadsheet models.
35:19And there are many assumptions in these
35:22models. So, because there are many
35:24assumptions but the logic works well,
35:28these models, these spreadsheet models
35:30are basically very complex estimation
35:34cases. So, your interviewer is going to
35:36be using a simple estimation case or
35:39sometimes a more complex estimation case
35:41because they want to see if you
35:44can model scenarios in consulting.
35:47You're also going to estimate things in
35:49meetings with clients and things like
35:52that but the core skill is can you
35:55model? And they're not going to test
35:57your Excel skills, they just want to
35:59test your reasoning and they can teach
36:01you Excel later.
36:03So, let's see some examples of this.
36:05What is the market size of road
36:07maintenance in the US?
36:10What are the revenues of a local
36:13Starbucks store? So, this you could if
36:16Starbucks was going to expand and they
36:18hired a consulting firm, they might want
36:20to ask this consulting firm to model
36:23the revenues of of stores in different
36:25parts of the country, for example.
36:28How much do pharma companies spend on
36:31R&D every year?
36:34So, these are much much harder than the
36:37usual how many golf balls would fit a
36:40Boeing 747
36:42but these questions are much more
36:44realistic
36:45compared to what you're going to find in
36:48your interview at McKinsey, Bain and BCG
36:50or any other top consulting firm. And
36:53they're also more realistic compared to
36:56what you're going to do on the job. So,
36:58you need to know how to answer these
37:00questions and there's a proper way to do
37:02that that you're going to learn in the
37:04estimation module which is the first
37:07module after this
37:09getting started module.
37:11The second analytical building block are
37:14quantitative analysis or put simply,
37:17analysis.
37:19So,
37:20when you have the data to calculate,
37:23when you have when you don't you have
37:24estimations. When you do have the data
37:26to calculate, we call it a quantitative
37:29analysis.
37:30And the thing to know here to be aware
37:32here is that getting to the right number
37:35is the minimum. How you do this and what
37:39you do after you do this
37:41is as important. Another thing to
37:44notice, even though you get the data
37:46when you have to do a quantitative
37:48analysis, you may not get all the data
37:52handed out to you. So, your interviewer
37:55might may hold some of it back to see if
37:57you even ask for it. So, a lot of people
38:00think that quantitative analysis in
38:02consulting is about your math skills and
38:05your quant skills but it's more about
38:08how you ask for data or if you even
38:11realize that you need to ask for the
38:13data and what you do with those numbers
38:16and with the answer.
38:18So, let's see some examples of this.
38:20Calculate the revenue synergies from
38:23introducing your competitors' products
38:26into Starbucks stores after an
38:29acquisition.
38:30So, you want to buy your competitor and
38:32one of the synergies is that you have
38:34access to the Starbucks shelves
38:38and you need to calculate these
38:39synergies. Now, calculating is the
38:43merely the beginning of it. You need to
38:45understand and make sense of what the
38:46results mean afterwards.
38:49Calculate cost savings in R&D after
38:52implementing this new process. And then
38:55the interviewer might hand you out a
38:57piece of paper
38:59with the new process.
39:01Given this table of data, how much will
39:04it cost to improve the quality of this
39:07university's dorm rooms? So, you're
39:10using numbers to reach conclusions. This
39:13is the essence of quantitative analysis.
39:18And finally, the last building block,
39:20chart interpretation. So, chart
39:23interpretation questions happen when the
39:26interviewers give you an already done
39:28analysis in the form of a chart or a
39:32table to see how you understand it and
39:34how you find the implications of it.
39:38So, great candidates in these types of
39:39questions, they find more insights. They
39:42communicate these insights better, and
39:45they develop better next steps.
39:48Remember when I told you that what you
39:51do after you get to the number in an
39:54analysis is the most important part?
39:57Well, there's a whole building block to
39:59it. It's the chart interpretation
40:01building block.
40:02Often, interviewers don't even want to
40:04check your math and your quant skills,
40:07or they've already checked earlier in
40:09the case, and they just want to see how
40:12much information, how much insight can
40:15you get from the numbers.
40:18And they'll give you a chart, which is a
40:21done analysis. It's finished, and they
40:23will tell you what can you tell me about
40:26the situation or about this client or
40:28about what they should do. And this
40:31resembles the higher level work of a
40:33partner. Partners don't crunch numbers.
40:36They don't build Excel models anymore,
40:39but they still look at the results of
40:41these models and of these analysis and
40:44help find the conclusions, the insights,
40:47and the next steps. So, whenever an
40:50interviewer gives you one or more
40:53charts, have this in mind. They're
40:55testing you
40:57for your potential to find these
40:59insights and next steps, and for your
41:01potential to become a partner someday.
41:04So, let's see some examples of this type
41:06of question.
41:07Maybe they give you a table comparing
41:09different countries that Nestle could
41:12invest on expansion and ask what do you
41:14think of this?
41:16Or maybe they give you a chart,
41:18a graph that charts the effectiveness
41:21and the addressable population and the
41:24cost of different public policies to
41:27reduce childhood obesity. And they ask
41:30you, well, what public policy would you
41:32choose and why?
41:34Or maybe they give you a chart showing
41:37the iron ore supply and demand curves
41:40with the prices and volumes of all the
41:42main global suppliers and buyers. And
41:45they ask you to draw conclusions from
41:48this supply and demand curve, and real
41:50supply and demand curves with real
41:52players and real prices, they look quite
41:55different from the ones that you might
41:57have learned in microeconomic classes.
42:01So, these are the types of questions
42:03that you can get in a chart
42:05interpretation question. Now, let me
42:07show you how you can solve any problem
42:11by combining these six building blocks
42:14together. And I'm going to show you that
42:16through two different examples of cases.
42:20And after watching those examples,
42:22you're going to see how the six building
42:24blocks fit together and how you can
42:27actually solve any type of case by using
42:30these six LEGO blocks.
42:33So, let me start showing you how to tie
42:35these six building blocks together in a
42:37an example of a product launch case. So,
42:40here's the case question. An engineering
42:43firm just developed a sensor that
42:46predicts mechanical problems in cars
42:49following a series of driving data that
42:52the sensor gets from the car itself.
42:55Should they launch it or not?
42:58So, if you get a case on a strategic
43:00decision such as this one, it's a
43:02decision, so the first thing you want to
43:04do is to lay out a plan to solve this
43:07problem and help the client make the
43:09decision. That means we should start
43:12with a framework.
43:14Something that breaks down the complex
43:15decision into simple problems. I'm going
43:19to go with a conceptual framework here
43:21because I think it's unrealistic to try
43:24to estimate the exact profits of this
43:27product launch. So,
43:29because this is a strategy question, I'm
43:31going to go into a more qualitative
43:34route. So, suppose I use a fairly simple
43:37framework with four categories.
43:40Is there going to be customer demand? Is
43:43there going to be a lot of competition
43:44in the future? How can we distribute
43:47this product efficiently? And is this
43:50product profitable early on? Is this
43:52product profitable after maturity? So,
43:55it's a framework. It's a simple one that
43:57I created from scratch, and of course,
44:01uh we also need specific questions and
44:05hypotheses for within each category. I'm
44:08not going to go through each and every
44:11question and hypothesis that I'd have in
44:13within each category because this would
44:15take too long, but suppose one of the
44:18issues within the first bucket is
44:21is there a customer segment that has a
44:25high desire for such a type of product
44:28and high willingness to pay uh for such
44:31a sensor.
44:33So, I want to know if they really want
44:35it and if they're willing to pay a lot
44:37for it.
44:38Let's say
44:40uh
44:42the interviewer tells me to start in
44:44this issue. So, if this is an
44:46interviewer-led case, they might ask,
44:48well, what do you think? They might say,
44:51well, let's start in this issue and and
44:53tell and say, what do you think are the
44:55main customer segments for this type of
44:58product?
45:00If this is a candidate-led case, if
45:02you're leading the case, you need to
45:05prioritize that question within your
45:07framework yourself. And then you need to
45:10be aware that to answer if there is such
45:13a customer segment with high desire and
45:15high willingness to pay, you also need a
45:18list of
45:20the customer segments because then you
45:22can see if
45:25if there is one of those customer
45:26segments or more that has high desire
45:29and high willingness to pay. Right? So,
45:32regardless if this is an interviewer-led
45:35case, you're going to reach this
45:38question. And
45:40what this means is you got to solve now
45:42a brainstorming problem. So, you have to
45:45create a list of hypotheses.
45:48In this case, hypothetical customer
45:50segments that might be wishing this type
45:52of product a lot and who are willing to
45:55pay for it. So, we started with the
45:57framework, and regardless if it's a
45:59candidate-led or interviewer-led
46:02interview, we reach a brainstorming
46:04question. Of course, there are other
46:06paths to take. This is just a
46:08hypothetical path that I'm laying out
46:10here. Now, if you're skilled in
46:13brainstorming, you'll know that you need
46:16to answer this question, what are the
46:18main types of customer segments
46:20for this type of product in a structured
46:22way. So, you might start saying that you
46:25can sell this to people who have cars or
46:28to companies that manufacture cars. And
46:31within the people that have cars, you
46:32can sell this to actual people, so the
46:34B2C market to consumers, or to companies
46:38that own cars. So, car fleets as home,
46:41B2B. So, I'm just giving you an example
46:43of a structure for this brainstorming
46:46question. Say you finish the
46:47brainstorming, you create these
46:49categories, and then you go on to
46:51specific ideas within each bucket. And
46:54one of those ideas is that large
46:56corporations with car fleets could be an
47:00ideal customer, right? They're rational
47:02in cost savings. They're corporations.
47:04They want to save costs, and they don't
47:07like to have their cars breaking down
47:09and stopping their operations.
47:12So, this is one idea that you have in
47:15your brainstorming. It's one hypothesis
47:17of a good customer segment. Uh in both a
47:21candidate-led and in an interviewer-led
47:24case, your interviewer might prompt you
47:26as you say that idea and say, well, how
47:28would you test
47:30the hypothesis
47:32>> [snorts]
47:32>> that this corporate car fleets are good
47:35customer segment for our client to
47:37launch this sensor? And this is a
47:40typical hypothesis testing question,
47:44which can be answered in a three-step
47:46approach that we teach in the hypothesis
47:48testing module. Right? So, you we
47:51started with the framework, a large list
47:53of questions and issues and hypotheses
47:56categorized into different categories to
47:58break down this complex decision into
48:00simpler problems. One of those problems
48:02was a brainstorming question, what are
48:04the main types of customer segments, and
48:06we found a hypothesis of one customer
48:08segment that could make sense.
48:11And now we need to test that hypothesis.
48:13And within your plan to test that
48:16hypothesis, something that might happen
48:19is that you might mention that you need
48:21to analyze how much cost savings will
48:25this new sensor bring per car in a
48:28corporate car fleet, right? And your
48:30interviewer might have data for that.
48:32So, if it's $10 per year per car in cost
48:36savings, companies are probably not
48:38going to buy your sensor, but if it's
48:40$500
48:42and you can charge much less for your
48:43sensor, then they you might have a
48:46market. You also might have mentioned
48:48that you need to estimate
48:51the number of existing corporate cars to
48:54see if this market is large.
48:56Right? And this is all to test your
48:59hypothesis. And your interviewer might
49:01say, well, do this estimation, do this
49:03analysis, and tell me what conclusions
49:06you get. And after you reach those
49:08conclusions, let's say that you found
49:10that there's quite a big number of
49:12corporate cars and that you can generate
49:15a lot of cost savings for them,
49:17they might give you a chart showing
49:20detailed data on corporate car usage and
49:23maintenance, so you can draw even more
49:26insight. So, maybe your product is good
49:28and there is a large market, but
49:30something about the usage or the the way
49:32these companies do maintenance
49:34invalidate your business module your
49:36business, uh, model, or make it even
49:40stronger case that you should launch
49:42this product. So, with these insights in
49:45hand, you may draw a partial conclusion
49:49on whether it makes sense to launch this
49:51product or not. And I say partial
49:54because
49:55there are other issues in your
49:57framework, uh, to be analyzed before you
49:59can categorically say, "Yes, you should
50:02launch this product." So, after you do
50:05this, you have to go back to your
50:07framework, see what other issues are
50:09there in the customer demand category,
50:11and in the other categories before you
50:13can make a recommendation.
50:16Now, not all interviews are this smooth.
50:20Of course, I I'm I simplified reality
50:22here, but your interviewer will guide
50:25you as you go, right? So, if you follow
50:28a path that won't lead nowhere, they
50:30won't give you data and have you waste
50:33time there. They'll just guide you and
50:35say, "Well, this is not really the
50:36interesting part. What What else do you
50:38think is most important?"
50:40They typically have some of these
50:42questions, some of these brainstorming,
50:45hypothesis testing, estimation,
50:47analysis, chart questions in mind before
50:50you get into the room, but they might
50:52create some on the spot, specifically
50:55estimations, brainstormings, and
50:57hypothesis testing. It's hard to create
51:00an analysis or a chart on the spot,
51:02although it happens. I've seen
51:04interviewers drawing charts on the spot.
51:07Uh, and the reason they create these
51:09questions on the spot is because you've
51:12said something that they never heard
51:14before, you had an interesting issue in
51:16your framework that they've never heard
51:17before, and they thought it was an
51:19interesting idea to explore that. So,
51:21some of these questions are going to be
51:23pre-planned, and some of them are going
51:25to come up on the spot, especially on
51:28candidate-led cases.
51:30Also, some interviewers, let's say you
51:32go through this whole process, you do
51:34the brainstorming, the hypothesis
51:36testing, the estimation, the analysis,
51:38and the chart interpretation that I
51:39mentioned, and, uh,
51:42some interviews will finish the case
51:44right there, even though you can't give
51:47a full answer to the case question,
51:49right?
51:50Uh, you you don't know if it's worth it
51:53or not to launch, you just know if
51:55there's
51:56uh, customer demand, or if at least
51:58there's a segment that makes sense to
52:00launch this product to. So, you can't
52:03really give a full recommendation to the
52:05case, but even then, uh, a lot of
52:08interviews will finish right here
52:09because your interviewer has already
52:11tested you on your reasoning skills.
52:15Other interviews will let you finish the
52:16case, they'll let you go through all the
52:18other issues in your framework,
52:21or some of the other most important
52:23issues in your framework, and they'll
52:25probably not spend a lot of time doing
52:27analysis and estimations, they'll just
52:30ask you how did you test it, give you
52:32the answer to the test, and you move on
52:34to the other issues.
52:36Both types of interviews work. Uh, those
52:40that let you finish the case and go
52:42through your whole framework, and those
52:44that will just test some building
52:45blocks. And they work because
52:48interviewers have tested your
52:50problem-solving skills through each
52:52building block. And honestly, they don't
52:55even have to test every single building
52:58block, as we're going to see in the next
53:00example. They have to test some of the
53:02building blocks, uh, so they they can
53:05see how you reason, how you think
53:07through the problem.
53:09Now, let's get a little, uh, let's step
53:12back from case interview world, and see
53:15how this would play out in a real
53:17project. In a real project, you will
53:19brainstorm a lot of hypotheses, way more
53:23than in a case interview. And you're
53:25going to test many of them, not all of
53:27them because you'll prioritize some, but
53:30you're going to test many of them.
53:32By doing several analyses, several
53:35estimations, or models, uh,
53:38uh,
53:38uh,
53:39scenario planning, and things like that.
53:42And you'll have to interpret all the
53:44results of all these estimations and
53:47analyses,
53:48uh, and also some qualitative data,
53:51uh, with charts that you're going to,
53:53uh, you're going to produce those charts
53:55to your manager, to your partner, to
53:57your client. So, that you can get a
54:00point of view, an understanding of this
54:02analysis in, uh, the
54:06in face of an overarching framework that
54:08you're going to be using to solve your
54:10client's problems. And you're going to
54:11keep adapting this framework as you find
54:14new things. So, this
54:17uh, scheme would be working for a
54:19project as well, not just for a case
54:21interview. Case interview is a
54:23simplified version of the real thing.
54:28So, the last example was of a more
54:30complex strategic case, and I used each
54:34building block once.
54:36But not all interviews are like that. In
54:39fact, the majority of them are not. It's
54:42common that a case doesn't have all the
54:44six building blocks, and it's also
54:46common that the same building block is
54:49used more than once in an interview. So,
54:51you can have two estimations in an
54:53interview, or three charts, or four
54:56brainstorming questions. So, let me show
54:59you a simpler case that's built from a
55:02different combination of the six
55:06building blocks. And here's the case
55:08question.
55:10A rental car company's maintenance cost.
55:13So, the maintenance costs of a company
55:15that rents cars
55:17have soared in the past year. What's
55:20happened, and what should they do?
55:23So, suppose for this case, I use a
55:25simple step-by-step framework because
55:28the question is what should they do,
55:31what decision should they take, I'm
55:32going to start with a framework again.
55:34And I'm going to start,
55:36uh, with a step-by-step one. First,
55:39identify why maintenance costs have
55:42increased. Second, brainstorm ideas, or
55:46generate ideas to solve the problem. And
55:48third, pick the best idea, pick the best
55:51solution.
55:53Now, obviously, I need a specific plan
55:56for each of the steps. I'm not going to
55:58lay it out here to, uh,
56:01to save you some time.
56:03But, the first thing we need is to start
56:07with an issue tree
56:08to diagnose the root cause of the
56:11problem. That's the first bucket. Why
56:13have maintenance costs increased? Well,
56:15the first step is to start with an issue
56:18tree, and I tell the interviewer that.
56:20So, an issue tree is nothing more than a
56:23MECE brainstorming of the drivers of
56:26maintenance costs. So, what are the
56:29drivers of maintenance costs? We need a
56:31MECE tree for that. Are these drivers
56:34increasing or decreasing? So, are
56:37maintenance costs increasing because
56:39we're using the cars more, or because
56:41the price per maintenance job has gone
56:43up, or because we're doing more
56:45maintenance jobs per car even though
56:48usage is flat?
56:51Could be all three, or could be also
56:53that we have more cars. So, basic
56:54structure.
56:56After you get this structure out, you
56:59need to get data for each driver. So,
57:04uh,
57:04you need to find out which driver is the
57:06most responsible for the rise. And this
57:09is a simple analysis, but it's an
57:12analysis nonetheless. So, it's another
57:14building block here. Say you found out
57:16that the problem comes from your unit
57:19maintenance costs getting higher. So,
57:22each time you send a car to do
57:23maintenance, you pay a larger bill.
57:26Now, you need to brainstorm hypotheses
57:29of why this is happening. You can't
57:33solve the problem without knowing the
57:35cause first. You can't solve the problem
57:37unless you know what are the possible
57:40reasons why that driver, the maintenance
57:43costs per unit of maintenance per per
57:46maintenance job, has increased. You need
57:48to know what changed.
57:50Now, assume you did this brainstorming,
57:52and you found that this is happening
57:54because you've internalized the
57:56maintenance operations, so it's more
57:58reliable, but also now your mechanics
58:01are not as productive because they spend
58:04a lot of time traveling around compared
58:06to the outsourced solution that you had
58:08before. Now, you could have needed a
58:11hypothesis testing building block to
58:13come to this conclusion, you could have
58:14needed to test this hypothesis, but
58:18often times the interviewer will just
58:20tell you that this is going on once you
58:22raise this hypothesis in the
58:25brainstorming. And this means you can
58:27jump straight to the brainstorming of
58:29potential solutions, which is the second
58:31part of our initial framework. So, now
58:34you found the the the why, you just move
58:36on to the second part.
58:38And once you find these ideas or
58:42potential solutions, uh, we need to pick
58:45the best solution. And we could pick
58:47that by analyzing which one of these
58:50solutions maximize our profit. So, we
58:53might want to ponder if there is,
58:56uh, if there are strategic reasons to
58:58choose one over the other, but we could
59:00also just go with the profit
59:02maximization route. And you have to
59:05choose if you're just doing profit
59:06maximization, or if you're considering
59:09other things as well. It depends on you,
59:11it depends on the case, it depends on
59:13the solutions you propose. And here,
59:15once you do this analysis, and you find
59:18out what the best solution is,
59:20the case is done, and you can just
59:21recommend,
59:23uh, what your client should do. So,
59:25we've gone through the whole framework,
59:28we've gone through the whole plan, and
59:30we have an answer as soon as we finish
59:32this last analysis. And
59:35this is a simpler case. It's very
59:37linear, and it's also a case that use
59:40just three building blocks. The
59:41framework, a couple or a few of
59:44brainstorming questions, and a couple of
59:46analysis questions.
59:49Now, before we finish this video, just
59:51some final thoughts. The first one is
59:54that there are some cases with just one
59:57building block. You've probably heard of
1:00:00market sizing cases where it's just the
1:00:03estimation case. You can have one
1:00:05estimation, which is the typical market
1:00:08sizing case, or you can have two
1:00:09estimations, three estimations within
1:00:12this case. There's also brainstorming
1:00:16only cases. It's very typical of final
1:00:19rounds at McKinsey, at Bain, and at BCG.
1:00:22Partners love it, and usually it happens
1:00:25when they will ask you to discuss the
1:00:27future of a new trend or of new
1:00:30industry. So, brainstorm how blockchain
1:00:34will impact the business model of banks.
1:00:37And they won't give you any data, any
1:00:39charts. You just have to brainstorm.
1:00:41They're very difficult cases, and you
1:00:44just solve them through several
1:00:46brainstormings
1:00:48uh chained together.
1:00:50There's also I've seen chart only cases.
1:00:52So, the interviewer might hand you
1:00:53three, four, five, six charts and ask
1:00:57you what's your conclusion to his
1:00:59client. So, you won't even structure the
1:01:01case. You just have to draw as many
1:01:03insights as you can from the charts, and
1:01:06of course tie them together. So, it's
1:01:08testing your structuring skills somehow,
1:01:11uh but only through charts. So, there
1:01:14are some cases with just one building
1:01:16block. And as there were as there are
1:01:18cases with all six building blocks and
1:01:21anywhere in between. Now, the second
1:01:24thought that I want you to leave this
1:01:26video with is that the difference
1:01:28between the candidate-led style and the
1:01:31interviewer-led style, which is the
1:01:34McKinsey style,
1:01:35is that when solving the candidate-led
1:01:38style, you are the person who decides
1:01:41which building block comes next. And
1:01:44that's the only difference between the
1:01:46two styles. This means that
1:01:48understanding when and why each building
1:01:52block is useful is the key to know how
1:01:55to guide your own cases. Some people
1:01:57have difficulty guiding their cases and
1:01:59knowing where to go next. And how do you
1:02:02do that? Well, you understand what's the
1:02:04function of each building block, and you
1:02:06understand that at the end of whatever
1:02:08you're doing, you need to know where to
1:02:10go next, to which building block to go
1:02:13next, as I've shown you in the examples.
1:02:16In the interviewer-led cases, in the
1:02:18McKinsey style cases, the interviewer
1:02:20will tell you what to do next. So, all
1:02:23you have to do is to solve each building
1:02:25block. But even then, it's useful
1:02:28knowing when and why each building block
1:02:30is useful, because as you finish solving
1:02:34one building block, if you show the
1:02:36interviewer and tell them what to do
1:02:38next, or in other words, what building
1:02:41block would you solve next, this is a
1:02:43huge differentiator from other
1:02:46candidates. Now, obviously, you don't
1:02:48have to you can't say, "Well, I finished
1:02:50this analysis, and now what I do is to
1:02:53use the brainstorming building block,
1:02:55right?" You have to say, "Well, I
1:02:57finished this analysis, and it indicates
1:02:59to me that the client should enter this
1:03:02market, and now I'd like to brainstorm
1:03:05uh what possible ways the clients can
1:03:09have to enter this market, so we can
1:03:12pick the best way." So, you're pointing
1:03:14to the direction that you want to go.
1:03:17You don't have you don't even have to
1:03:18use the word brainstorming. And the last
1:03:21thing I want to I want to leave you with
1:03:24are the benefits of this approach. And I
1:03:27know we've talked about them in the
1:03:29beginning of the of this video, but I
1:03:31want you to have them very clear in your
1:03:34mind, because now you understand how the
1:03:36approach works. The first benefit is
1:03:39that it's easy to self-diagnose where
1:03:42you should improve. It's easy for you
1:03:46to know where you should improve when
1:03:49you use this approach, because you have
1:03:51to improve either in one specific
1:03:53building block or in guiding the case.
1:03:56That means choosing which one comes
1:03:58next. Second benefit, it's easier to
1:04:02practice by yourself. You can do drills
1:04:05for each building block, and you're
1:04:07going to see how to do that through this
1:04:09course. So, it's much like sports
1:04:12people, they practice shooting,
1:04:14dribbling, passing the ball. You can do
1:04:16drills for each building block.
1:04:19Next thing, there's more clarity on what
1:04:22to do to improve. So, say you
1:04:24self-diagnose that you need to improve
1:04:26uh your hypothesis testing.
1:04:29With other approaches, you're you're
1:04:31left wondering, "Well, how can I improve
1:04:34hypothesis testing?" But now, there's a
1:04:37lot more clarity. Each building block
1:04:40has a step-by-step method. It's easy to
1:04:43see in which step of your solution
1:04:46you're messing up. And then, we're going
1:04:48to show you some techniques for each
1:04:50step. So, say you're
1:04:53having trouble with step number two of
1:04:55hypothesis testing. Then you can see
1:04:57what techniques work for step number
1:04:59two, and there's way more clarity on
1:05:03what to do to improve. It's way more
1:05:06practical.
1:05:07Fourth, there's more clarity on how to
1:05:10deal with weird or unusual cases.
1:05:14You can stop and think when you start a
1:05:16case like this. Which building block
1:05:18should I use first? And then second? And
1:05:21then third? And if you do that, this
1:05:23will help you create a step-by-step plan
1:05:26to tackle the case, which is exactly
1:05:28what the interviewers want to see in an
1:05:31unusual case. Right? And finally, you
1:05:34always know how to answer any question
1:05:37if you use the six building blocks.
1:05:40Interviewers always want you to either
1:05:43structure problems or draw insights from
1:05:47data. Usually numerical, sometimes
1:05:49qualitative as well. And it's usually
1:05:52really obvious which building block
1:05:54their question is related to. And once
1:05:56you know which building block is their
1:05:59question related to, you can use the
1:06:01right technique to answer their
1:06:03question. And if you use the right
1:06:05technique, the right step-by-step
1:06:06method, you check all their boxes. So,
1:06:10you always know how to answer any
1:06:12question using this method.
1:06:14Right? I hope you've enjoyed. And the
1:06:17next steps are to watch the next two
1:06:19videos in this module to understand the
1:06:21tools that you use through your
1:06:24preparation and the mindset that you
1:06:26should have. The next video on the tools
1:06:29is very tactical. So, what can you do to
1:06:33improve certain things? And the one
1:06:35following that is very mindset-oriented,
1:06:38but they're both really, really
1:06:40important. Then, after you do that, move
1:06:43on to the estimation module, and then
1:06:45after that, the brainstorming module, so
1:06:47that you can start learning your first
1:06:50two building blocks. I hope you've liked
1:06:53this video. I know it was kind of long.
1:06:55You're probably a bit tired right now,
1:06:57uh but I hope you've liked it, and I
1:06:59hope it gave you a new perspective on
1:07:01case interviews. And if it did, shoot me
1:07:05an email. I'd love to hear about it.
1:07:07Okay? Thank you so much. See you in the
1:07:10next video.

About this transcript

This page contains the full transcript of Case Interview Fundamentals: Lesson 1 [The 6 Building Blocks of Case Interviews] by CraftingCases, generated from the public captions YouTube serves with the video. The transcript has 10,364 words across 1,605 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.