Prompt interpretation

Hey there,

Sure! Let me go question by question

1. Are there some tips and tricks I can use before creating a structure and how do I come up with initial hypotheses to test in order to get a more relevant structure?

My suggestion for clarifying questions and subsequent hypotheses is the following:

• Start with 1-2 converging questions and make sure you have: (i) a tangible, quantifiable objective/target, (ii) an indication about constraints
  --> focus is on defining the solution space as narrowly as possible (going from broad objective to narrow solution scope)
• Then, ask 1-2 few diverging questions to understand better the context (going from no context to some context)

The last diverging questions are the point where you can try to gather more information to build hypothesis. Here, there's no fixed rule: it all depends on your ability to read the prompt and understand the company/industry. 

Let me give you an example: if it's a steel foundry with a profitability issue, you should focus more on quantity and fixed cost than price and variable cost: this industry has vast scale economies and capex efficiencies and variable cost are standardized commodities. So, a good diverging question would be "I would like to understand whether our volume size and growth has been in line with our competitors, I know this is a sector that enjoys massive scale economies. Are we a larger (but inefficient) market leader or a mid-market company that might suffer from lack of scale?"

2. How do I check if my math approach is simple and how to ensure it's structured?

If you are a fast talker, I suggest you use a 3-step approach (see full example below from one of my custom "drills" document):

• Align on the logic -- and confirm the approach
• Set up the equation with variables -- and confirm the approach
• Plug in the numbers -- and comment the final result

Interviewers don't like fast talkers who never check-in with them. This approach gives you room to talk and lead the case, while at the same time giving you a chance to make this more conversational

3. What should I do in the 2-4 seconds before speaking? I tend to be a fast talker :)

Most MBB folks are fast talker. The problem might be that you just dive into the exercise/maths/brainstorming without clarifying:

• Context (5 seconds)
• Objective (5 seconds)
• Plan (15-20 seconds)
• And only then executing

Clients always want us to share a clear plan, rather than throwing out numbers. That is what interviewers want you to mimick!

Hope this helps, and good luck on your interview :)

Best,

Tom

___

Hi Victor, 

The feedback you received is actually pretty encouraging. If BCG told you your structure was "spot on", I would be careful not to over-correct and create a new problem before Round 2.

On your questions:

1. Hypotheses before the structure

Personally, I wouldn't force yourself to generate 2-3 hypotheses for every case.

A lot of candidates hear "hypothesis-driven" and think they need to come up with a brilliant answer in the first 30 seconds. That's not really the point.

Usually I try to answer two simple questions:

• What is the objective?
• What are the most likely drivers of that objective?

For example, if profits are declining, I don't need a sophisticated hypothesis. A perfectly reasonable starting point is: "My initial hypothesis is that either revenues have deteriorated or costs have increased, and I'd like to understand which effect is more significant."

The structure then helps me test that hypothesis.

The goal is not to be right. The goal is to have a starting point.

2. Simpler math

The biggest mistake I see is candidates trying to be clever.

Before touching the numbers, ask yourself: "What is the simplest path from the data I have to the answer I need?"

Often there are multiple ways to calculate something. The best one is usually the one with:

• fewer steps
• rounder numbers
• less mental load

When practicing, don't just check whether your answer is correct. Ask yourself: "Could I have solved this in half the steps?"

That's often where the improvement comes from.

3. The 2-4 seconds before speaking

Use them.

Seriously.

Many candidates think silence feels awkward. Interviewers generally don't.

One habit that helped me a lot was forcing myself to mentally answer: "What is the one thing I want to say?"

before opening my mouth.

You'd be surprised how much cleaner your communication becomes.

In fact, based on the feedback you received, if I were you I'd spend more time practicing:

• slowing down
• taking pauses
• simplifying math

than trying to reinvent your structuring approach.

For a Senior Associate Round 2, clear thinking and mature communication will probably move the needle more than finding the perfect initial hypothesis.

Your R1 feedback is genuinely good. Strong structure is the hardest part to fix and you already have that. The rest is mechanical and doable in a week.

On hypotheses, don't try to think and structure at the same time. Sequence it. Take 20 to 30 seconds to ingest the prompt, then ask 3 to 4 clarifying questions right away. These buy you time and surface hypotheses naturally. While listening to the answers, mentally form 2 to 3 working hypotheses. Then structure for 60 to 90 seconds.

On math, always say the approach out loud in plain English before touching numbers. Round aggressively, 1,000 is easier than 1,047. Walk the interviewer through your setup before computing, they'll nudge you if you're off.

On pausing, this matters a lot for fast talkers. Just breathe before every answer. Pick your headline first, then your supporting points, then speak. The pause feels long to you but lands totally fine to them.

Good luck.

hey Victor, 

For the prompt, don’t force hypotheses. Just restate the goal in your own words, pull out the one or two drivers that obviously matter, and turn those into simple clarifying questions. That’s enough. Your mind goes blank because you’re trying to be clever, aim for simple and directional, not perfect.

For math, the test is: can you explain your plan in one sentence before calculating. If you can say “I’ll take revenue, divide by customers, then adjust for mix,” it’s simple enough. Keep every step visible and verbal.

In the 2–4 seconds before speaking, breathe, look at your notes, and decide on your first sentence only. That’s all you need to slow down and sound controlled.

Best, Alessa

Victor, it seems you've already received some great answers. 

The questions you're asking are great, but they require a deeper diagnostic and analysis. I would actually need to see you approaching a case to be able to give you valuable feedback and explain how to practice in order to close the performance gap. 

If you decide you'd like some help, feel free to drop me a line. 

Best,
Cristian

You shouldn't be trying to solve the case before you come up with an approach. So it is actually good that you are not being able to come up with those 2-3 hypothesis before the structure - because it will look like you are "fishing" for an answer, and that signals a non-structured candidate.

So keep up doing what you are doing right now, as it is working and from a conceptual POV it is better