"I couldn't believe these 'easy' market sizing questions cost me the offer at THREE top consulting firms"
Back in 2012, when I first applied to jobs in consulting, I made a frustrating, costly mistake...
I underestimated estimation and market sizing questions.
And then, they bit me in the ass.
Not once... Not twice...
I failed estimations in THREE DIFFERENT FIRMS that year.
(Over four different cases... Two different McKinsey partners tested me on this.)
You can see it for yourself:
I don't know what's worse: to fuck up two estimations in the only final round that I got to go to (McKinsey) or to miss the opportunity to EVEN GO TO A FINAL ROUND at two other target firms because of a couple of (seemingly) straightforward market sizing cases...
Back then, I thought estimations were supposed to be easy. I didn't give them enough credit.
I had spent all my time working on my analyses (making sure I wouldn't make any math mistakes), and trying to guarantee I'd be able to structure any business and public sector cases that came my way...
You know, doing those things that most candidates spend most of their time on.
And then, by the time I went to my actual interviews...
I was overconfident and underprepared when it came to estimations
And you know what?
I wish I could tell you it was just bad luck... That the interviewers had a bad day, or that I got an unusually difficult estimation...
But I can't. It wasn't bad luck.
I got my first warning sign when I was rejected at Kearney by a market sizing case...
Then at BCG...
But I still ignored them when it was time to go to my McKinsey Final Rounds:
"Partners will obviously test me with something tougher than mere estimation questions..."
That's what I thought, despite my experience in the first rounds at the other firms.
And then the McKinsey partners did exactly the opposite of what I was expecting.
I got TWO estimation questions in my McKinsey final rounds.
And they weren't your usual market sizing case... They were MUCH HARDER.
I hadn't even prepared enough for "regular" market sizing cases...
Imagine what happened when I got these "partner-level estimations" that I had to solve in front of two McKinsey partners?
Of course, I blundered both of those interviews and threw my last chance away.
One of the partners told me in the feedback call: "We really liked you... But we don't think you're ready. I encourage you to try again next year..."
Yep, I'd have to wait a full YEAR before I could interview again.
This simple, silly mistake of underestimating estimations has cost me three offers.
The months of preparation, dozens of mock-interviews and several rounds of real interviews with interviewers who had (otherwise) liked me were now completely irrelevant.
All of that time and effort WASTED.
And this didn't just happen to me...
I wish I could say I'm an outlier.
I wish I could tell you that very few candidates have trouble with estimation questions, or that these indeed are the "easy type of questions" to most candidates.
But the truth is that candidates get rejected due to simple estimation questions ALL THE TIME:
B2B estimations are TOUGH. Now, think about it... This guy's WHOLE CAREER would've been different had he been proficient in solving this type of estimation case when he had that interview.
And it's not just TOUGH estimation cases that get candidates rejected... Take a look at this guy:
This solution isn't too bad on paper, but candidates often do slight mistakes on the way they justify their reasoning, communicate their approach and react to follow-up questions that put in check their credibility with their interviewers.
This of course, begs the question...
Why does this happen?
Why do candidates get rejected in estimations and market sizing cases so often, in first and final rounds?
The reason is in plain sight...
No one teaches you how to answer estimation questions properly
You and I know that there's a ton of subpar preparation material out there...
But I can't think of a single topic that is taught in a more ATROCIOUS way than estimations and market sizing cases.
Take a look at this one paragraph from a popular book:
Why on earth would someone think this way of categorizing estimations would be useful to any candidate?
It's like categorizing healthy food items into Fruits, Vegetables and PREPOSTEROUSLY HEALTHY FOODS.
I could make fun of Mr. Cosentino all day, but I'd be missing the target...
All the other resources I've ever seen have the same three fatal problems that will severely hamper your chances of getting the job if (when) you get an estimation question in your interviews...
FATAL PROBLEM #1: The examples they give are NOTHING like the real questions you'll get in real interviews
Imagine you were studying for a test. Let's say it's a math test...
One of the first things you'd look for are examples of questions (and answers) that resemble the ones you'll get in the exam. You'd look for questions similar in type and difficulty.
Well, guess what?
The questions you see floating around in books and websites are way less complex and much, much easier than the ones you'll find in real interviews -- especially at McKinsey, BCG, and Bain.
(And firms like Kearney, Roland Berger, and Strategy& do often have some extra tricky estimation cases.)
For instance, here are the examples that a popular website mentions as representative of consulting estimation questions:
No wonder some candidates skip preparing for estimations properly...
They think this kind of stuff is the real deal!!
Even Youtube doesn't do a much better job...
Not long ago, Julio set off to watch every video on estimations and market sizing out there and curate the best ones for you in a new article in our website.
Here's the email he sent after his saga:
It's borderline IMPOSSIBLE to learn estimations properly if all you see are examples that are on the easy side.
It's like going to a math test thinking you'll get one-variable equations and getting differential calculus instead.
You can study as much as you want, you'll still fail miserably.
But having examples that are way too easy is only the first fatal problem of current resources...
FATAL PROBLEM #2: Current resources give you examples of answers that WOULD NOT PASS REAL INTERVIEWS.
This one gets me REAL MAD.
For instance, take a look at this proposed answer to an estimation question in Case in Point:
Your interviewers would challenge these assumptions very strongly ("How do you know there are just 6 stations in your town?", "Why is your town representative of the whole US?", "What makes you so sure of the 5,000 figure?", "What about trucks and highways?") and you'd fail the interview if not able to address these concerns.
Even if we overlook the fact that this guy's suggesting you to make shit up because it divides by 30 (???), and try to look at the bright side, to the fact that his structure at least is mathematically right...
... Still, 80-90% of interviewers would reject a candidate had he "solved" this case like this.
To me, a good preparation resource is one that teaches you how to answer questions in a way that ACTUALLY GETS YOU OFFERS.
Anything less than that is subpar.
And it's not just Case in Point...
Coincidentally, I found on another popular website another answer example to the same "gas station" estimation (which goes a long way to show how much variety current resources are showing you...)
Here, take a look at it:
Now, this is definitely a much better way to estimate the # of gas stations in a given location.
But would it pass the interview?
And here's why: as important as having a logical tree and some assumptions hanging on top of it is...
... It's just as important how you'd defend your logic and assumptions to an interviewer.
Two different candidates could have completely different results (e.g. getting the offer vs. being dinged before the interview was even over) using the exact same tree and the exact same assumptions.
Why? Because they'd justify their structures and assumptions in different ways... Then one of them would reality check the answer well enough to make the interviewer comfortable and the other would not.
A good answer has things like:
- The logic through which you'd defend the specific structure you're using to an interviewer (vs. an alternative approach).
- Why the specific level of depth they're suggesting in the answer was chosen.
- The reasoning behind EACH assumption (which this last example leaves you wondering, for example) -- this is SUPER important, and interviewers will grill you on this all the time in a real interview.
- The specific segmentation pattern that would make the most sense to certain assumptions (and those that wouldn't make sense at all).
- The reality checking comparison that would be used at the end to validate that the number you got to makes sense or not.
Having those things talked through in your answer (vs. simply a structure and some numbers) makes it complete. It puts the interviewer at ease that you're someone who'll be able to model scenarios in a reasonable way once you join their teams, and that you actually know what you're doing.
It's very hard to learn to solve estimations well from current resources without those kinds of nuances... Even worse when they give you answers that are plainly wrong.
(Especially because, as a candidate, you can't always differentiate the good from the bad.)
But there's a third fatal problem that makes it almost impossible to learn estimations well enough to feel confident and competent in your interviews...
FATAL PROBLEM #3: There's no proven method that GUARANTEES you'll solve any estimation and market sizing case.
Suppose you're really driven to really learn to solve estimations at a high level of performance.
You really want to get the offer, after all.
Here's what you're handed as instructions...
"Base your assumptions on some sort of logic".
Preposterous advice, straight from a famous book.
... Then I'm sorry to bring you the bad news, but you're pretty much SCREWED.
I'm not even saying some of this stuff is plainly wrong... I've already said that.
The main problem is that it is NOT EVEN PRACTICAL.
Pretty much all advice on how to solve market sizing cases out there boils down to the following: be logical and use reasonable assumptions.
Yeah -- duh.
I guess you could say that to race car drivers as well: "When racing, make logical decisions, based on reasonable assumptions."
Or to theoretical physicists: "When trying to disprove string theory, use logic and good assumptions."
The problem with that kind of advice is that, even though it sounds good, it doesn't actually help you when you have real struggles, such as:
- Your interviewer just asked you 'what's the market size for professional cameras in your country' and you're there frozen, nervous as hell trying to come up with an approach that will work and not make your life harder...
- Your interviewer doesn't like the one structure you came up with and now you have to come up with yet another one, as if the first wasn't hard enough already...
- You're pressed to justify an assumption that, honestly, you just said a number without giving much thought to it... And now you're stuck in your head trying to find a reason why it makes sense but all the reasons feel like BS to you...
Unless you want to risk failing at three firms because of estimations as I did, you'll need something more reliable than that.
You'll need to know the patterns and systems that allow you to know exactly how to approach any estimation question an interviewer gives you instantly, no matter how weird or difficult they are...
You'll need proven techniques for structuring, segmenting and assumptions that consistently delight interviewers... And then, you'll need methods to reality-check your results with confidence so you can finish your interviews on a high note.
You'll also need preparation material that warns you about the common traps interviewers will put you in, and teaches you how to safely get out of those.
No current material on the market does that. They simply say "Hey, all you need is to be logical -- trust me, everything's gonna be okay".
So, nowadays we have resources that teach estimation questions and market sizing cases that:
- Are way too easy and much simpler than the ones you're gonna get
- Show you unreliable (and sometimes plainly wrong) answers to those questions
- Don't give you an actual method to solve these questions that you can rely on
No wonder so many candidates are caught unaware and get rejected with estimation cases.
Could it get any worse?
Estimations are WAY MORE COMMON than you think
I used to think of estimation cases as an afterthought.
I thought they were mainly used by boutique companies or early at first rounds.
This was due to both ignorance and the impression that I got from the very same materials that didn't teach me how to solve them well.
I had the perception (from these materials) that estimations weren't that important, and I guess some part of me thought that if my "trusted materials" weren't giving it enough importance, it must be because they didn't happen often in real interviews.
I spilled my water and slipped on it...
So, how common are estimations?
Julio and I have coached hundreds of candidates through thousands of interviews at different firms and offices, across all continents.
Here's what we've found:
- You'll get an estimation question in about 40-50% of the first-round interviews you have, except at McKinsey. McKinsey very rarely has estimation questions in their first-round cases (though that could happen).
- You'll get an estimation question in about 30-40% of final-round (partner) interviews, including at McKinsey.
And here are the nuances:
- In first rounds at BCG and Bain, the estimations you get are likely to be 5-10 minute questions within broader cases (market entry, private equity, etc.).
- In other firms' first rounds, you're as likely to get an estimation question within a case, as you are to get a pure estimation case (i.e. the only case question you get in the whole interview is an estimation, usually market sizing). In a pure estimation scenario, there's very little room for error as they have no other way to evaluate you.
- When partners do estimations (again, 30-40% chance), they're either the only question they ask or the main question they ask (and they'll ask other questions after you do the estimation). Partners also tend to have very challenging estimations that are often unusual.
Still don't think it's a big deal?
Well, if you consider that you have a 30% chance of getting an estimation in every interview, what do you think are the chances that you'll get at least one estimation in each firm you apply?
It turns out that if you have a 30% chance of getting an estimation per interview, and you have 6 interviews at a firm, you have a whopping 88% chance of getting at least one estimation before you get an offer at that firm.
Check the math out:
A 30% chance of getting an estimation is a 70% chance that you won't.
70%^6 gives you the probability you won't get an estimation in 6 interviews: 12%...
... Which means there's an 88% chance you'll get one in at least one of those interviews
You can run your own assumptions, and the conclusion is pretty much always the same:
You're almost sure to get at least one estimation question before you get an offer from a consulting firm.
(And often you'll get two or more of these questions, just like it happened to me in my McKinsey final round.)
And when you do get one of these questions, you better be pretty damn sure about what you're doing, because...
Estimations are HIGH STAKES -- they can get you rejected more easily than any other type of question
Make a math mistake in a normal Quantitative Analysis question, and you get a ton of opportunities to recover: these questions are always one small question within a much broader case...
Forget a key factor or two in a Framework or Brainstorming question and you're usually fine: if you had a bunch of other good key insights and just forgot one or two small things, your interviewer will usually forgive you for that...
Estimations are different. They're high stakes.
Unlike pure structuring questions (Frameworks, Brainstormings), if you forget one single variable, your whole structure is wrong -- the equation doesn't add up.
It doesn't really matter how many other insights you had. An equation without all variables is a wrong equation, and a wrong equation can cost you the offer.
The other thing that makes them high stakes is that, unlike pure analytical questions (Analyses, Charts), there's little room for recovery.
Making a mistake in an Analysis, for example, can be pretty bad, but you have a couple things going for you... You're usually 10-15 minutes into the case, so you built credibility with your interviewer already. Also, you'll usually have other analytical questions in the case, so you can prove your mistake was an isolated incident.
Not with estimations.
In the scenario where the estimation you got is the only question of the case -- if it is the case itself -- you can't mess it up for obvious reasons.
There won't be another question where you can make it up for your mistakes. The whole impression the interviewer will have of you will be dependent on that.
Now, not always the estimation will be the standalone question of the case...
But even when it is not, it is usually a KEY question.
If you get an estimation within a market entry, product launch or private equity case, for instance, that is usually one of the most important questions of the case, and one that will take a good chunk of the time too.
In both of these situations, messing up the estimation is messing up the case.
And, in most firms, if you mess up the case, you've messed up the round.
You might be invited for another interview, but most likely that will be, as it happened to me, a year from now.
But there is a third factor that raises the stakes for estimations:
They're the one single type of question that can give your interviewer a well-rounded assessment of your consulting skills.
Specifically, that's why partners love estimation questions so much.
Estimations are sort of a "mini-case".
They're much simpler than a full-fledged business case, but they test your structuring, business sense, analytical and communication skills altogether.
Estimations are a simple, but complete way for an interviewer trying to assess your fundamentals. Each part of the estimation process reveals key skills to the interviewer -- and they can always follow-up with more questions to learn more about how you think.
So, to sum up, estimation and market sizing questions are:
- High stakes. And thus, highly likely to be a deciding factor in your offer should you get one in an interview...
- Frequent. You're almost certain to get in any firm you apply to, even the top ones...
- Largely ignored. Traditional materials don't give you realistic examples, reliable answers and a trustworthy method to solve them. And because of that, most candidates are underprepared.
No wonder so many otherwise well-prepared, intelligent candidates fail due to this type of question so much.
The good news: it doesn't have to be this way
Fast forward to the year after I was rejected. I decided to follow that partner's advice and apply again to all consulting firms I could.
So, what happened? Did I get estimations in my interviews this second time I applied?
Of course I did.
But this time, things were different.
I nailed every single estimation questions that I got.
Not only that, there was no luck involved. I knew I'd nail each and every one of them as soon as the interviewer spoke the words -- before I even put pen to paper.
For instance, my last interview at Bain was with a senior partner. Not any senior partner, but the office manager.
Yep, the top guy.
What did he choose to interview me with?
You guessed it... An estimation case.
Here's the case question:
Estimate BMW's call center costs.
As soon as he spoke those words, I smiled.
Even though it's not your usual estimation question, I knew I could effortlessly build a customized structure for it.
I did it, and presented a 3-layer structure to him.
"Good job, but I think we'll need to go deeper than that."
As soon I broke down each bucket of the structure, he'd ask me to go one layer further.
It ended up being a 50-minute estimation case.
Whenever I structured a bucket, the partner would incessantly ask for a deeper and deeper structure.
When it came time for choosing the assumptions, he didn't go light on me either -- he doubled down on questioning me and challenged every assumption as if he did that for a living (in a way, he did).
Can you imagine being nervous, in front of the most senior partner in the office, having to go deeper and deeper in your rationale in what seems like a never-ending quest for a more detailed MECE structure and better thought-out assumptions and not losing your nerves?
He'd ask things like:
- "Have you considered that these call center reps will take vacations?"
- "What about our employees that get pregnant? Doesn't that add to the cost?"
- "Shouldn't the time delay before the rep picks up the phone affect your cost per minute?"
- "How do you think the fact that BMW is a premium brand affects how often their customers actually call their call center? And how would you know how long these call are? Your personal experience isn't a good basis for these assumptions..."
I had never seen an estimation case go this way.
He wanted so much detail into my structure that I ended up with TWO pieces of paper just to fit the structure.
I pulled it off and ended up with a 6-7 layer MECE structure and the most thought out assumptions of my life.
Julio and I included this Bain Final Round estimation as a drill in our estimations course, Master Market Sizing
As I finished the case, exhausted but proud of the work I did, he smiled broadly:
"Very good job, I think you're one of the best people I've ever seen at this"
At that moment I knew I'd get that offer.
But regardless of the offer, I just felt happy with myself.
I was proud that I went from losing three offers because of estimations to really feeling confident at solving one of the toughest ones that were ever thrown at me.
I later talked to other candidates who were being interviewed by the same partner that day. They all said he used the same case and was as detail-oriented and picky with them as well.
Many were a bit bitter about the whole situation -- "I wish I had prepared for something like this, but how on earth would I know this was a possibility?"
I didn't talk to them again, but my guess from how they were feeling is that most of them got rejected that day.
I, unlike the other candidates, didn't feel bitter, nor anxious.
I solved the case, I didn't get stuck, I had strong answers to all his follow-up questions and the partner himself said he really liked me.
And later that day, I got the offer.
What was different this time around?
The true difference between this time and the first time I applied to consulting and got multiple rejections was that in the meantime, I stopped behaving like a "candidate" and started behaving like a real consultant.
What made my interviewer like me so much was that, as I was solving the case, he felt like he was talking to a fellow Bainee, not just another candidate.
But how exactly did I achieve that?
Between the first and the second time I applied to consulting, one thing happened: I worked for 9 months in Venture Capital.
This would be completely irrelevant to this story if it weren't for two random facts:
1) In Venture Capital you have to estimate markets, revenues and even individual cost lines of a start-up all the time.
2) I worked directly with two former consultants.
Fast forward a few months, and while every other candidate was trying to impress that Bain partner "using logic and assumptions", as Case in Point and other resources suggest... I was solving that exactly as his own consultants did in the real project that inspired that case.
I had an unfair advantage...
All I had to do to solve that estimation and to answer any follow-up question that came my way was to do exactly what I did at the VC firm day in and day out.
What's more, because I directly worked with two former consultants, the methods I was using to estimate in the VC firm, as well as in that partner interview, were the exact same methods that the consultants that partner worked every day with used to estimate things for their own clients.
This gave me the ability to:
- Instantly know 2-3 approaches to any estimation the interviewer threw at me, and which one was best in which situations...
- "Read my interviewer's mind" whenever he asked a follow-up question to know exactly why he was asking that question (my former boss would ask me the exact same questions), and then give the perfect response to address exactly his concerns...
- Know which methods would feel reasonable to my interviewer whenever choosing an assumption or reality checking a result -- and which ones would feel like complete BS (I had 9 months of experience being called out on the BS).
Now, the bad news is that it doesn't make sense for you to get a job in Venture Capital and spend 9 months learning this skillset like I did...
The good news is that you can learn the exact same skillset in just 1-2 weeks -- without having to switch jobs.
... and now you can have an unfair advantage too.
Last August, Julio and I were sitting at a coffee shop discussing much of what I'm telling you today.
We were talking about how poorly prepared most candidates are when it comes to estimation cases...
... And how that's expected, as the resources that they have to prepare for this type of question are lousy and misleading.
To put it nicely, we weren't happy with what was out there.
(I was outrageously annoyed, irritated, PISSED OFF that the bar for content quality was so low.)
We were also talking about how many people were emailing us, asking for us to make more content on estimations and market sizing:
And then came the thought...
What if we took everything we know about estimations and market sizing and made a comprehensive course that helps candidates become outstanding in solving them?
We took out a notebook, ordered another coffee and started brainstorming ideas of what such a course should have.
Initially, we had HUNDREDS of ideas.
But as we dug deeper, we noticed that there were only 3 things candidates really needed:
Patterns, techniques and practice.
- Patterns allow you to hear an estimation question from an interviewer and instantly know at least 2-3 proven approaches to solve it, even if it's a question you've never heard before. They take all the risk off the table, ensuring you can answer any question, no matter how easy, hard or obscure.
- Techniques help candidates with specific weaknesses and challenges. Some candidates have a real hard time creating a mathematical structure, coming up with defensible assumptions, or even doing the math. Yet, you're always one technique away from turning a major weakness into an invaluable strength.
- Practice helps you internalize the patterns and the techniques. Focused practice in a wide variety of estimation cases makes you feel confident and prepared to face any question. It gives you the experience you need to answer questions like a real consultant.
At the end of the day, we were over-caffeinated and eager to create that course.
We spent hundreds of hours cracking the code to estimations for you
Over the next few months, we rolled up our sleeves and designed our dream course on estimations.
We created a list of all estimation and market sizing questions we saw in real interviews.
These were cases that Julio and I got personally while interviewing at firms like McKinsey, BCG, Bain, Kearney, Roland Berger, Monitor Deloitte, etc.
Then, we added cases that former coaching clients got at dozens of different firms and hundreds of different offices.
We wanted to make sure that the method we created was going to work for all of them.
We also insisted that the examples we had in the course were realistic, not oversimplified questions like the ones you see in traditional materials.
We then created our system: a combination of patterns that made solving any estimation case easy.
This ended up becoming what we now call "The Estimation Blueprint".
Think of it as an "algorithm" to notice the pattern behind the estimation question that is asked, and to solve it flawlessly.
Finally, we gathered our best, proven techniques that have helped real candidates overcome common challenges in the estimation process.
- Could teach people who had trouble with structuring, how to structure any estimation with ease.
- Gave "instant" creativity to non-creative candidates on how to come up with no-BS assumptions and reality checks.
- Made math quick and error-free even to people who couldn't do math in their heads.
- Allowed candidates to quickly come up with insightful answers to your interviewers' most common follow-up questions.
In the end, we had a system like no other...
This system will help you answer ANY estimation and market sizing question (even the most obscure ones)
After creating this system, we reality checked it against the list of market sizing problems we had from real interviews.
It was easy to use this system to solve every single one of them, flawlessly.
Yes, even the most obscure questions. Even those that 90% of candidates wouldn't even know how to start. Even those filled with trick questions and misleading information designed to derail all but the best candidates out there.
And the reason why this was possible is simple:
This system is nothing more than a collection of the patterns and techniques that real consultants use when estimating numbers.
Just like I learned to estimate markets, revenue and cost figures, and other numbers like consultants do when I was working in Venture Capital with former consultants... You can learn the same skillset (though much, much faster) by learning this system we put together.
And because the patterns and techniques are exactly the same that these consultants use, you'll be able to estimate the same numbers that they themselves can, using the same methods that they themselves use (though most are unconscious that this is what they're doing).
It's the shortest path to:
- Effortlessly solve any estimation question that an interviewer throws at you.
- Feel confident you can get yourself out of any situation, be it a tricky assumption that's hard to justify or a sneaky follow-up question that you've never heard of before.
- And have your interviewer let out a sigh of relief as you start solving the case, as they see that they've finally found a candidate that "gets it" and knows what they're doing.
And today, you're one of the first to find out about it...