Case Interview Fundamentals: Lesson 36 [Drill: Drugstore chain inventory losses] — Transcript
Full transcript
- 0:01Hey, everyone. How's it going? This is
- 0:02Julio speaking again, and let's start
- 0:05chart interpretation drills.
- 0:07Just so you know one thing about this
- 0:09drill, the goal here is that you
- 0:11practice almost exclusively coming up
- 0:14with conclusions based on the data, and
- 0:17then coming up with next steps.
- 0:20These charts that I'm going that we're
- 0:21going to use in this drill are simple on
- 0:23purpose, but they are still realistic.
- 0:26If this was If this were a real case
- 0:29interview, what the interviewer would be
- 0:31testing is a lot more your capacity to
- 0:35draw conclusions and come up with
- 0:38high-quality next steps based on the
- 0:41case that we're doing, rather than
- 0:42whether you can read the chart or not.
- 0:44So, sometimes the charts are going to be
- 0:46simple to read, and of course, the only
- 0:47way to stand out in those cases is
- 0:50coming up with amazing conclusions and
- 0:53next steps. So, practice for this in
- 0:55this drill.
- 0:57Now, as you know from the previous
- 0:58video, you need at least a simple
- 1:01framework for the case, so that you can
- 1:03draw conclusions from your chart. But if
- 1:05you're starting out, taking time in
- 1:07creating a very complete and deep
- 1:09framework will help you get more quality
- 1:11practice in the chart interpretation
- 1:14question as well. So, I do recommend you
- 1:16stop and build a high-quality framework.
- 1:19I am going to show you my own here, but
- 1:21it's always good practice to create
- 1:23yours. Okay? Now, let's move on to the
- 1:25case question.
- 1:27Your client is a medium-sized drugstore
- 1:30chain that has been facing inventory
- 1:33losses.
- 1:34This problem has been there for as long
- 1:35as they have been tracking inventory
- 1:38losses data. Every year, their inventory
- 1:40losses amount to 1.5% of revenues in
- 1:45dollar costs,
- 1:46while they know the expected benchmark
- 1:48is 0.6%.
- 1:50They've hired you to help them bring
- 1:52this metric down to the benchmark. And
- 1:55the case question is, how would you
- 1:57approach this problem? Now, pause this
- 1:59video and come up with the best
- 2:02framework that you can. I'm going to
- 2:03show you my framework afterwards, and
- 2:05then you can use either mine or yours.
- 2:09Preferably yours. This is going to be a
- 2:10lot better practice to use your own.
- 2:12But, if there's something missing in
- 2:14yours that mine has, I'd recommend you
- 2:16adapt yours so that we have so that you
- 2:19have all of the important issues in
- 2:21sight.
- 2:22The way I'd approach this case is I
- 2:25would try to first find exactly and
- 2:28precisely what the problem is that is
- 2:30causing this
- 2:32above benchmark level of losses or of
- 2:35inventory losses.
- 2:37Then, based on the very specific
- 2:40problem that I found, I would brainstorm
- 2:42solutions and then pick the best
- 2:45solutions either with best ROI or most
- 2:48efficacy, depending on what the client
- 2:51wants. Now, I want to dig deeper into
- 2:54how I would find this problem and what
- 2:57level of specificity I mean when I talk
- 3:00about finding the problem. So,
- 3:01basically,
- 3:03this problem is within the company, and
- 3:06so I assume that I can
- 3:09find a very specific source of the
- 3:13problem. And the way I would do this is
- 3:14first, I'd break down these inventory
- 3:17losses, if I have the data, and I assume
- 3:19that I do, by
- 3:22first, what types of losses are we
- 3:24talking about? Are we talking about
- 3:25damage or expiration date or something
- 3:29else?
- 3:30Second, I would try to find where in the
- 3:32value chain they are happening. So, is
- 3:34it happening in receiving the products?
- 3:37Is it happening in the distribution
- 3:38center? Is it happening in putting the
- 3:41products in the trucks or during
- 3:43transport or in placing them in stores?
- 3:46You know, where is it happening?
- 3:48Then, I try and also I try to find
- 3:51what types of products we are losing.
- 3:53So, I could find a pattern in there. Are
- 3:56we losing mostly medicine? Are we losing
- 3:59mostly beauty products? Are we losing
- 4:01mostly edibles? And then I also try to
- 4:04find
- 4:05a geographic breakdown so that I can
- 4:07tell if this is happening in a specific
- 4:10group of stores or if it's not, which is
- 4:13also useful information.
- 4:15And with these breakdowns, I would
- 4:17expect to find things like let's say
- 4:20these very specific transport routes are
- 4:24experiencing inventory losses with
- 4:26shampoos due to damage. And so the
- 4:29causes could be things like the
- 4:30transport firm we are using
- 4:33is not careful with this type of
- 4:35product. Or the tools that they use, for
- 4:37example, the box where the products go
- 4:39in
- 4:40or the way they use these tools. So,
- 4:43maybe in this area
- 4:45the the boxes need to be thrown on the
- 4:47ground or whatever for checks or for
- 4:50unboxing
- 4:51end up damaging the product. Or maybe
- 4:53the way that they stack the
- 4:56the boxes one on top of each other end
- 4:59up damaging this product. And maybe they
- 5:01just need new processes. All right.
- 5:03Another example of things that I could
- 5:05find edibles end up being lost in the
- 5:08stores, all of the stores, all around
- 5:11our geography because they are past
- 5:14their due date. And we might find out
- 5:16that this is because
- 5:18we purchase edibles in larger bulks
- 5:21because that's also how we purchase
- 5:23everything else. And other items have
- 5:26larger shelf life, but the edibles
- 5:28don't. So, we might need to change the
- 5:30purchasing product process.
- 5:32Or I might find out that on specific
- 5:34locations uh
- 5:36OTC meds, over-the-counter medicine, are
- 5:39disappearing at alarming rates. And it
- 5:42could even be organized theft because,
- 5:45you know, some gang might have found out
- 5:47a way around our security systems or
- 5:49whatever.
- 5:50And after finding after doing these data
- 5:54breakdowns and finding a specific
- 5:57numerical location for the losses,
- 6:00then I would further investigate on the
- 6:04hypothesis that I can have. So, I'll
- 6:06come up with ideas like the ones I just
- 6:08told you for why this might be happening
- 6:11and then I'm going to
- 6:13go see whatever is happening, go talk to
- 6:16whoever is responsible because in most
- 6:19of the cases there's probably going to
- 6:20be a responsible person for the whatever
- 6:23processes is causing or seeing the
- 6:26losses so that I can find a real-life
- 6:30reason for these losses to be happening.
- 6:33And then, I'd be to brainstorm
- 6:35solutions, but that's only that
- 6:37completely depends on what the causes
- 6:39are that we find out. All right?
- 6:43And this is the framework I would show
- 6:46my interviewer. This is how I present it
- 6:48to him. And now, let's move on to the
- 6:51next question, the question we're all
- 6:53looking forward to, the
- 6:55chart interpretation question.
- 6:57All right. Now, let's say that the
- 6:59interviewer showed me this exhibit and
- 7:01they asked me what do I think. Right?
- 7:04Now, pause the video and come up with
- 7:06your answer. Go through the four-step
- 7:09approach thoroughly and then watch my
- 7:12answer, okay? I'm going to guide you
- 7:13through all of the steps here even
- 7:15though I would only say to the
- 7:17interviewer the fourth step, all right?
- 7:20So, step number one is mapping the
- 7:21territory. Basically, I want to ask
- 7:23myself
- 7:24what this chart can tell me given the
- 7:27data that it's showing me without even
- 7:29looking at the data, right? And
- 7:30basically, what it can tell me is
- 7:33whether this is an expiration problem if
- 7:36we're above benchmark in expiration
- 7:39losses or it's a damage problem if we're
- 7:42above the benchmark in
- 7:45damage losses or if it's both if we're
- 7:47above the benchmark on both. Of course,
- 7:50we've got to be above the benchmark in
- 7:52one of them because in the aggregate we
- 7:54are above the benchmark.
- 7:56And I would guess that
- 7:58since missing products aren't there,
- 8:01then if we're not above the benchmark in
- 8:02either, I'd have to brainstorm
- 8:05what would cause us to be above the
- 8:07benchmark if not those two and I my
- 8:09guess would be it could be missing
- 8:11products like products just disappear.
- 8:13Right? Now, let's move on to step number
- 8:15two and read the nuance. And basically
- 8:18the only thing to do in this step is
- 8:21read the footnote and understand that
- 8:23it's saying that the data comes from
- 8:25client data and from the consulting
- 8:27firm's own research. What this means is
- 8:29basically I can trust the data. Don't
- 8:31worry about that. All right. Now, step
- 8:34number three is reading the data and
- 8:36since I've done step one well,
- 8:39this is going to be easy and it's
- 8:41basically telling me that we are
- 8:431% above the benchmark in damage losses
- 8:47and so that's probably the cause for
- 8:50the losses that we're experiencing.
- 8:52Actually, that's definitely the cause
- 8:54for the losses we're experiencing cuz it
- 8:56responds for even more than the gap that
- 8:59we're trying to close given the case
- 9:00question. All right. Now, let's move on
- 9:03to step number four that's interpreting
- 9:05the data. So, as I said in the last
- 9:07video, I want to have the framework next
- 9:10to my
- 9:11insights from the chart and I want to
- 9:13connect them. And so basically
- 9:16this is telling me this chart is telling
- 9:18me what type of loss we are dealing with
- 9:20and the answer is damage.
- 9:23Now, what I should do is choose what to
- 9:25do next and why that's my choice and how
- 9:29I would do it. So, is it to dig more to
- 9:31find out more about the problem or is it
- 9:34to solve the problem? And whichever of
- 9:36those I pick, how so? Right? Right now,
- 9:40in my interpretation, I still don't know
- 9:42enough about the problem in the level of
- 9:44detail that I need in order to
- 9:46brainstorm solutions. So, my choice is
- 9:48to dig deeper. And that's basically what
- 9:50my framework is telling me as well cuz I
- 9:52asked for all of these segmentations of
- 9:55our losses, and I only have one right
- 9:57now. So, that's what my framework would
- 9:59tell me
- 10:00if I basically looked at it. Now, I
- 10:03should choose a few of those options to
- 10:06go next. And so, here's what I would say
- 10:07to the interviewer. That next
- 10:10after understanding that damage is the
- 10:14predominant kind of loss that is causing
- 10:17all of the difference that we're seeing
- 10:18that we're trying to solve, I would try
- 10:20to understand where in the value chain
- 10:23this damage is happening so that I can
- 10:26understand exactly what's going on, and
- 10:28perhaps I can even find someone to call
- 10:31and discuss the issue, someone that's
- 10:32responsible for the process that's
- 10:34causing the loss.
- 10:35Before I do that, though, I would also
- 10:37want to look at what type of product is
- 10:40being damaged since the damage is
- 10:42probably a combination of the tools, the
- 10:45handling, and the product or package
- 10:48design. So, I need to understand what
- 10:51product are being damaged and where it's
- 10:53happening. All right? And now let's say
- 10:56that the interviewer gives me yet
- 10:57another chart and asks me what I think.
- 11:03Now, let's basically repeat the same
- 11:05process. Pause the video, come up with
- 11:07your answer in the four steps, and then
- 11:09listen to mine.
- 11:10So, basically, step number one,
- 11:12re-mapping the territory, is fairly
- 11:15simple again.
- 11:17This chart can easily tell me whether
- 11:19the damage problem is concentrated
- 11:21mostly in the reception of products, in
- 11:24the distribution center, in transport,
- 11:27or in the stores. And after that, I will
- 11:30have super clear next steps because I'll
- 11:32know exactly where to look for the
- 11:35problem. I might even be able to go see
- 11:38the problem happening in real life.
- 11:41And it can also tell me if the problem
- 11:44is not concentrated at all. And if
- 11:46that's the case, then it'd be a much
- 11:47more fuzzy problem and I would hope to
- 11:50remain exploratory and get insights from
- 11:53the other segmentations that I have
- 11:56rather than talking to someone right
- 11:58away. All right? Now,
- 12:00next step is reading the nuance and
- 12:03simply it's telling me again that the
- 12:06data is trustworthy. Now, let's go to
- 12:08reading the data. What this chart is
- 12:10telling me, based on what I can see and
- 12:12based on what I
- 12:14evaluated in step number one, is that
- 12:17most of the damage is happening in the
- 12:20distribution centers, while most of the
- 12:22expiration is happening in the stores.
- 12:25Since expiration is below the benchmark,
- 12:27I'm going to ignore this. It's normal,
- 12:30probably, and I'm going to focus in the
- 12:33damage happening in the distribution
- 12:34center.
- 12:35Now, for step number four, I want to
- 12:37have my framework with me again. And
- 12:39what I can tell is that the step in the
- 12:43value chain where the losses the excess
- 12:45losses are happening is in distribution
- 12:47centers. And of course, as I know as
- 12:49I've known from the previous chart, they
- 12:52are be products being damaged.
- 12:55All right? Before I can brainstorm
- 12:57solutions, I still need to know
- 13:00why they're being damaged and how, so
- 13:02that I can see what I need to change
- 13:05actually. All right? And this this
- 13:07framework gives me again a ton of
- 13:09options to choose where to dig.
- 13:11Basically, what I would want to see
- 13:14right now is a breakdown of the losses
- 13:16by distribution centers, so that I can
- 13:19understand which distribution center
- 13:21it's happening in and then I can
- 13:23hopefully talk to someone.
- 13:25It might even be a good idea to get a
- 13:27sector breakdown so that I can see what
- 13:30sector in the distribution center or in
- 13:32the distribution centers
- 13:34the damage is happening the most and
- 13:37then hopefully I can get one or several
- 13:39operators on the phone and talk to them
- 13:41about what's going on and what processes
- 13:43they're using and perhaps go see them
- 13:45which is even better than talking to
- 13:47just a distribution center manager cuz
- 13:49they're the they're the people that are
- 13:51actually doing the job and looking at
- 13:53other people doing the job. All right. I
- 13:56would also as I said before want to see
- 13:59a breakdown by type of product that's
- 14:01being damaged and if possible by type of
- 14:03damage. I'm not sure if we're recording
- 14:05this but we might be we might be able to
- 14:08tell right away if the products are
- 14:11being damaged meaning they are getting
- 14:14wrinkled or wet or they're getting
- 14:17yellow or whatever. All right. And so
- 14:19this is what I want to see before I can
- 14:21do anything else.
- 14:23Now let's say that interviewer pauses me
- 14:25right here again and says that is all
- 14:26very nice. Take a look at these exhibits
- 14:29and tell me what you think.
- 14:32Now let's ask ourselves what can this
- 14:36chart tell us, right? It's showing me
- 14:38the breakdown of damage losses by
- 14:41distribution center and by region and
- 14:44this chart can tell me if these losses
- 14:46are concentrated in a specific
- 14:48distribution center or not and if so in
- 14:51which distribution center and the same
- 14:53for regions. It can tell me if it's
- 14:54concentrated in one single region or not
- 14:57and if so in which region. But since one
- 15:01region can have more than one
- 15:02distribution center but each
- 15:04distribution center can only be in one
- 15:06region cuz it can only be in one place.
- 15:08The real
- 15:10breakdown here like the the right
- 15:13breakdown to look at it's distribution
- 15:14center. So honestly
- 15:17it's unless I look at these data this
- 15:19data and I see that it is not
- 15:22concentrated in any distribution center,
- 15:24but it is concentrated in a region
- 15:26specific region. And that would be
- 15:28weird, but that would be super
- 15:30interesting to look at.
- 15:31Then I'm not going to care about region.
- 15:33Cuz if it is concentrated in
- 15:34distribution center and region, which of
- 15:36course you can see that it is,
- 15:38then it's probably just a distribution
- 15:41center problem. Cuz
- 15:43a distribution center is only in one
- 15:45region. So if it's focused in this
- 15:47distribution center, it's going to be
- 15:48focused in this one region.
- 15:51All right. So basically this can tell me
- 15:53which distribution center is having the
- 15:55most problems, is having the majority of
- 15:57the problem, if any. And
- 16:00later we're going to see that it is in
- 16:02one specific distribution center.
- 16:04All right. And there isn't much to
- 16:07understand about the
- 16:09the nuance, right? But we can see
- 16:12and when it comes to data, we can see
- 16:14that the
- 16:17problem is clearly concentrated in
- 16:19distribution center number three. Now,
- 16:21let's get our framework out and connect
- 16:24it to the framework. So first thing I
- 16:26can see that
- 16:29if
- 16:30we can bring distribution center number
- 16:32three down to
- 16:35the benchmark in terms of the
- 16:38volume of damage losses that we have in
- 16:39it, we will have solved 100% of the
- 16:44inventory losses problem, right? Now, I
- 16:47still need to know exactly why and how
- 16:50the damage is happening. But now that
- 16:52I've pinned the location, I can go to
- 16:54distribution center number three and see
- 16:57the process. I can talk to whoever's
- 16:59responsible. It could be the
- 17:00distribution center manager or the
- 17:02sector the responsible for whoever's
- 17:04responsible for the sector. But before I
- 17:06talk to them, I would want to see, as
- 17:08I've said before, a sector breakdown of
- 17:11those losses so that I can understand
- 17:14where in the distribution center they
- 17:16are happening, a breakdown of the losses
- 17:18by type of product and by type of
- 17:20damage, as I've said before. And then, I
- 17:23can have more hypotheses of what might
- 17:25be happening. But still, I want to talk
- 17:28to the operator if possible, or the
- 17:31distribution center manager if not
- 17:33possible talking to the operator. And I
- 17:35would ask them, number one, were they
- 17:37aware of this problem? Do they know that
- 17:39it's happening? Number two, why do they
- 17:41think this might be happening?
- 17:43Number three, who is responsible for
- 17:46that very specific cause in the damage?
- 17:48And number four, what's the process
- 17:51like? What is the that specific handling
- 17:54of that specific product? What does it
- 17:56look like? And I'd ask the other
- 17:58distribution centers the same question
- 18:00because they are not having problems.
- 18:02So,
- 18:03if they describe the a very same
- 18:05process,
- 18:07then it could be that either
- 18:09distribution center number three is
- 18:10doing the process in a wrong way, or
- 18:12they don't have the best tools, or their
- 18:15people are not properly trained, or just
- 18:18not good enough employees. Maybe it
- 18:20could be a recruiting problem, who
- 18:22knows. Right? So, that is exactly what
- 18:24my next step would be.
- 18:26Okay?
- 18:27Right guys, this is it for this one
- 18:30drill. Now, I want to talk about just a
- 18:32few lessons that I want you to take home
- 18:34from this drill.
- 18:38Lesson number one, and if this feels
- 18:40repetitive to you, great, then you're
- 18:42capturing the message. If this doesn't,
- 18:45then you really need to understand this.
- 18:47When you're creating your next steps,
- 18:49you should guide your team or your
- 18:51client towards exactly what they need to
- 18:54do further. And when I say exactly, and
- 18:56this is what I want you to take from
- 18:57this
- 18:58third exhibit, my answer for the third
- 19:00exhibit is
- 19:02really precisely what you would do. If
- 19:05you're going to investigate, how exactly
- 19:07are you going to investigate? What data
- 19:08are you going to look at? What are you
- 19:09going to do with this data? If you're
- 19:11going to talk to someone,
- 19:13what are you going to ask them? Why are
- 19:15you going to talk to them? If you're
- 19:16going to act, if you're going to do
- 19:18something, what exactly are you going to
- 19:20do and how are you going to do it?
- 19:22So, what you want to show to the
- 19:23interviewer is that if you're the one to
- 19:26build the next deck of flights, you know
- 19:28exactly how you would build it. If
- 19:30you're the one to decide the next course
- 19:33of action that the client needs to take
- 19:35to solve this problem, you are the
- 19:37person that they want solving this
- 19:39problem. Okay?
- 19:41Mhm.
- 19:42Uh lesson number two here is when coming
- 19:45up with next steps, your framework is
- 19:47your greatest ally. You're going to see
- 19:50this more in the next drill. You also
- 19:52you could also see it in this one. Once
- 19:54I understand where this piece of data
- 19:57goes in my framework,
- 19:59it's obvious what I need to do next
- 20:01because I've already outlined what I
- 20:03need to do in the whole project. So, if
- 20:05I have this piece of data, it's the
- 20:07everything that's around it that I need
- 20:09to do next. I can either, as I said in
- 20:11the previous video, finish the current
- 20:13branch, I could
- 20:15move on to the next branch, or I could
- 20:17even create an entirely new branch. But,
- 20:20it's something within the framework
- 20:22because the framework is made to solving
- 20:24the case. And this this piece of data
- 20:27that you just got is also made to help
- 20:30you solve the case. Okay, so
- 20:32have your framework in front of you,
- 20:34especially in practice, and draw lines
- 20:36between
- 20:37your conclusions on the chart and your
- 20:40framework so that you can come up with
- 20:42insights and understand exactly how they
- 20:44help you solve the case. Okay?
- 20:47Lesson number three,
- 20:49and this is something that didn't happen
- 20:51in this case, but that I want to how I
- 20:53want to outline it to you here so that
- 20:55you won't
- 20:56not learn it in this case. And that is
- 20:59quantify your answers. In this case, it
- 21:02seems like we couldn't quantify too much
- 21:05because I never said, "Oh, 40% of this,
- 21:0960% of that, you know, $2 million
- 21:13but the thing is it's in this case most
- 21:15of the conclusions were
- 21:17basically 100%. So, I was actually
- 21:20quantifying when I was saying all of the
- 21:23damage losses happen in distribution
- 21:25centers. All of the difference between
- 21:27us and the benchmark in terms of
- 21:29inventory losses happened
- 21:32within the damage category.
- 21:35All of that was my way of quantifying in
- 21:37this case where most of the conclusions
- 21:40were 100%
- 21:42right? But it's still
- 21:44key to know that quantifying your
- 21:46answers is going to set yourself apart
- 21:49from different candidates. So, keep this
- 21:52in mind not only for the next drill but
- 21:54also for every time you get a chart
- 21:57interpretation question in mock
- 21:58interviews or in
- 22:01real life case interviews.
About this transcript
This page contains the full transcript of Case Interview Fundamentals: Lesson 36 [Drill: Drugstore chain inventory losses] by CraftingCases, generated from the public captions YouTube serves with the video. The transcript has 3,703 words across 571 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.