Struggling on McKinsey Math Questions

Hey! 

Totally relatable struggle, and the good news is that McKinsey case math isn't actually hard math. It's basic arithmetic done quickly under pressure, which is a very different skill. At the same time, math fluency is close to a precondition for getting through an MBB interview. You can have a brilliant structure and great communication, but if you stall on a 15% of 4B calculation, the interviewer notices. The bar isn't genius level. It's fast and confident on basics.

Here are the 4 main things to drill for McKinsey math. I tell my coachees that if they master these, then 90% of case calculations become trivial:

1. Percentages via the anchor method. 

• Memorize only 5 anchors (1%, 5%, 10%, 25%, 50%) and derive everything else by adding or multiplying them
• These anchors are very easy to find: e.g., 10% of X is 0.X
• 15% = 10% + 5%. 16% = 10% + 5% + 1%. 12.5% = half of 25%
• Also drill the reverse: "240 is 15% of what?" becomes 240 ÷ 15 × 100 = 1,600. This shows up constantly in profitability and market share questions.

2. Percent increase and decrease. 

This is where most candidates lose points, because the traps are conceptual, not arithmetic. Growth, margin changes, and price/volume questions all live here. Three things to internalize: 

• +20% then minus 20% does not return you to the start (you end at 96, not 100)
• Chained percentages multiply, they don't add, so two consecutive +10% growth years are +21%, not +20%;
• Reverse problems like "after a 25% discount, the price is 450, what was the original?" become 450 ÷ 0.75 = 600 (not 450 + 25%). 

3. The four operations with shortcuts. 

Stop doing long multiplication, a few examples:

• Round and correct for addition/subtraction (487 + 326 becomes 500 + 326 minus 13).
• Use the "fraction of 100" trick for multiplications by 25, 50, 75, 12.5 (48 × 75 = 48 × 3 ÷ 4 × 100 = 3,600).
• Decompose for awkward multiplications (17 × 14 = 17 × 10 plus 17 × 4 = 238).
• Halve three times for divisions by 8.

4. k/M/B notation, treating suffixes like exponents. 

• Never write zeros in a case.
• Handle the number and the suffix separately: 50M × 200 becomes (50 × 200) M = 10,000M = 10B.
• Memorize the rules: k × k = M, k × M = B, B ÷ M = k.
• Combine with percentages for market sizing: 15% of 4B = 0.6B = 600M.

How to actually practice

10 minutes a day, timed, max 10 seconds per calculation, written by hand. Consistency beats long sessions. Drill one of the 4 areas per session rather than mixing, because you build pattern recognition faster that way. Once individual drills feel easy, simulate the real cognitive load: do mental math while pacing the room, or while explaining your structure out loud. That's the actual interview condition. 

One last thing on public maths (i.e., maths in front of someone else) 

When you freeze in the moment, it's almost always because you didn't set up the calculation before starting it. Take 5 seconds to say out loud, "OK, I need to find X, which equals Y times Z." Interviewers love this because it shows structure and it gives your brain a clear path. Diving into numbers without setup is what causes the blank.

For you and for everyone else: DM me if you want a PDF document with some drills!

Best,

Tom

Hi there,

You’re already on the right track, my biggest recommendation is to keep doing exactly what you’re doing. Some of the most valuable learning in case prep actually comes from the review process itself.

A strategy I strongly recommend is to revisit each case about 24 hours after completing it. Redo the case on your own, then review it in detail and take structured notes on your performance, key takeaways, and areas for improvement.

This approach helps you recognize recurring patterns in case types, sharpen your problem-solving framework, and improve how you structure your thinking over time. The more you reflect on past cases, the easier it becomes to apply those lessons across future ones.

Hope that helps!

Annika

Hi,

This is very normal — especially with McKinsey-style math.

What’s happening is not that the math is hard. It’s that in the moment, you’re not sure how to set it up, so you hesitate. Once you see the solution, the logic looks obvious.

A few things that usually help:

First, don’t rush into calculations.
Take a few seconds and ask yourself: what exactly am I trying to find?
Then link it to something simple like:

• revenue = price × volume
• profit = revenue – cost

Most questions reduce to one of these. If you get the setup right, the math becomes much easier.

Second, get used to the common patterns.
McKinsey repeats similar types:

• % changes (growth, decline, margins)
• breakeven
• weighted averages
• quick market sizing

If you’ve seen them enough times, you start recognizing them quickly instead of figuring them out from scratch.

Third, say your steps out loud.
Something like: “I’ll first calculate X, then Y, then combine them.”
It helps you stay structured and buys you a bit of thinking time.

Fourth, simplify the numbers.
You’re allowed to round. Turning 49 into 50 or 98 into 100 makes a big difference, and interviewers don’t care about perfect precision.

Finally, practice math separately.
Even 10–15 minutes a day helps — percentages, ratios, breakeven setups. Also redo old case math without looking at the solution.

If I had to put it simply:
it’s not a math problem, it’s a setup problem.
Once you get into the habit of structuring first, those “easy in hindsight” questions start feeling easy in the moment too.

Best,

Soheil

Hi, this is very common, what you’re describing is basically the gap between “I understand it” and “I can do it under pressure.”

A few things that usually help:

First, slow down when you get the question. Most people struggle because they rush into calculations without being clear on what they’re solving. Before touching numbers, force yourself to say (even out loud):
“Ok, I need to calculate X, which depends on A and B.”

Second, break it into steps. McKinsey math is rarely complex — it’s just layered. Instead of trying to do everything at once, do:

• step 1
• check
• step 2

You’ll make fewer mistakes and feel less overwhelmed.

Third, write everything down. Trying to hold numbers in your head is where things fall apart in interviews.

On practice: don’t just do full cases. Do short drills.
Take 10–15 mins and practice only:

• percentages
• growth rates
• simple multiplications/divisions

Also take case questions and solve just the math part, without the full case context.

In terms of question types, yes — patterns repeat a lot:

• % changes (increase/decrease)
• breakeven / profitability
• market sizing with simple math
• weighted averages

If you get comfortable with these, you cover most of what comes up.

Last point: the reason it looks “easy” afterward is because you see the path. In the interview, your job is just to create that path step by step. That’s a skill you build with repetition.

If you want, happy to run through a few of these together. I’s one of those things that clicks pretty fast once you structure it properly.

Hi there, 

Sorry to hear about this struggle. 

For what it's worth it, it's absolutely normal as part of learning to solve cases. To a certain extent, it will get better.

With the candidates that I've worked, I noticed that most of them struggle with calculation questions because they don't have the right technique to approach them. 

Specifically, they tend to do too many things at once: communicating with the interviewer, developing the logic / approach, doing the computations, etc.

The technique that I teach aims to separate these stages and ensure that you co-create the approach and answer with the interview (just the same way you'd do this in consulting with the client, which is why interviewers appreciate this approach so much). 

If you need help with this, drop me a line.

Best,
Cristian

Are there specific question types that frequently come up - NO ! McKinsey math (or for that matter consulting math) is built on the same core skills (percentages, growth rates, weighted averages, breakeven, ROI) but the wrapping changes every time.

Two things that help. First, find people who are genuinely good at case math and run through the logic with them, you will pick up tips and tricks much faster. Second, do standalone math drills on the platform, some of them explain the reasoning well and that builds your intuition over time.

The fact that solutions feel obvious in hindsight is a good sign. It means the gap is pattern recognition, which only comes with reps.