Case Interview Math Practice

Why Most People Waste Weeks on the Wrong Kind of Math Drill

The standard advice for case interview math is terrible. You'll find endless posts telling you to grind 100 GMAT problems or memorize every formula in existence. That's not how this works, and anyone who tried it can tell you it doesn't translate to interview performance. Case interview math practice is fundamentally different from academic or standardized test math. It's about speed, approximation, and mental stamina under pressure—not precision or proof-writing. I spent about three months doing what everyone told me to do first. I was grinding quant sections from the Official Guide, getting perfect scores on those problems, and then bombing every mock case I sat for because I froze on basic percentage calculations mid-interview. The gap between test math and case math is enormous. Test math gives you clean numbers and unlimited time. Case math throws 47.3% at you while you're simultaneously structuring a market sizing problem and trying to read the interviewer's face.

Case Interview Math Practice

The effective approach is much more specific than generic math drills. You need to practice mental arithmetic in the exact context where it matters: under time pressure, with messy numbers, while holding multiple streams of thought in your head simultaneously. The single most useful exercise is called chunking and killing. You take a complex calculation like 15% of $4.7 million and break it into 10% ($470K), 5% ($235K), and add them. You practice this until it's automatic. Most candidates I've seen who actually improve this way spend 20 to 30 minutes a day on dedicated mental math sets, not hours of generalized problem solving. There are specific number patterns that appear constantly and most people don't prepare for them. One-third of 49. The square root of 50 times 7. These show up in market sizing and profitability cases so frequently that you should have them memorized. I keep a mental list of roughly two dozen common fractions and their decimal equivalents—1/7 is 14.3%, 3/8 is 37.5%, things like that. When I'm practicing, I write out 10 random business scenarios with embedded calculations and time myself. If I'm getting a result in under 45 seconds and it's in the right ballpark, I move on. If it takes longer or I second-guess myself, I write down exactly where I stumbled. Here's a concrete example of something I ran into during a practice session that changed how I approach these problems. I was working through a profitability decline case where the interviewer gave me a revenue figure of $23.8 million and a cost structure that included a variable cost of 62% and a fixed cost that grew by 8.4% year over year. I immediately went to calculate the fixed cost change as 8.4% of whatever base figure, and I completely missed that the interviewer hadn't actually given me the prior year's fixed cost number yet. I'd committed to a calculation path before having enough information. I lost about three minutes recalculating and the whole case trajectory shifted. After that, I started deliberately pausing after each data point in practice sessions instead of rushing into computation. It felt slow at first but it actually saved time in the long run because I stopped making unnecessary calculations on incomplete inputs.

Some of the more useful techniques that candidates overlook involve simplifying the arithmetic before you commit to it. Rounding 4.7 to 5 when it's a multiplier changes your answer by less than 10% in most business contexts, and the interviewer is usually more interested in whether your logic is sound than whether you got the third decimal place right. The real test isn't accuracy—it's whether you can communicate your calculation steps clearly while doing them. I've seen strong candidates fail cases simply because they went silent for 90 seconds while calculating mentally instead of narrating their process. Even saying "I'm going to round this to 5 for simplicity" out loud buys you time and shows structured thinking. The tools themselves matter less than consistency. There are apps like Math for Business and Fakebook that generate practice problems, and there are free question banks from case prep companies. I found that the timed question sets from major consulting firm career pages were actually the most realistic because the numbers had that slightly awkward quality real interviewers use. Something like "calculate the NPV assuming a discount rate of 11.3%" feels artificial but it's exactly the kind of thing you'll encounter. Here's something most guides won't tell you: the ceiling of your math ability in a case interview is usually determined by stress management, not skill. When I was doing my mock cases in the final weeks before interviews, I noticed a clear pattern. My arithmetic was fine when I was calm. The moment the case got structurally confusing—when I wasn't sure what question I was actually answering—the same calculations that took me 20 seconds would take two minutes and I'd make silly errors. This isn't a math problem. It's a working memory problem. Under stress, your brain has fewer resources available for calculation. The workaround is to practice doing mental math while intentionally introducing distractions. I'd have people talk to me while I worked through profitability problems, or I'd stand up and walk around while calculating. It sounds ridiculous but it made a measurable difference in my interview performance because the environment during actual cases is inherently distracting.

There are also limits to what you can realistically achieve. You cannot become genuinely fast at all types of calculations. Division with non-round numbers will always be slower than percentage calculations, and if a case heavily requires compound interest or amortization formulas, you're better off stating the formula clearly and asking for a calculator rather than attempting it mentally. Some interviewers will offer a calculator. Some won't. Know which firms typically provide one and prepare accordingly. McKinsey and BCG generally don't. Deloitte and PwC often do. This detail matters more than people realize. One more thing that isn't widely discussed: the relationship between your math speed and your ability to spot when an answer is wrong. This is actually more valuable than raw speed. After hundreds of practice problems, you develop an intuition for whether a result is plausible. If you calculate that a $2 million market grows to $8 million in three years at 15% growth and it takes you five minutes to get there, something is probably wrong. A quick sanity check—15% of 2 is 3, so year one alone adds 300K, meaning the answer should be closer to 2.8 million—tells you immediately that you made an error. Building this instinct takes time but it's built through repetition of the same problem types, not through variety. For a structured practice routine, I'd recommend starting with 15 minutes of pure arithmetic drills daily, moving to 30-minute timed case sections where you focus on a single quantitative component like breakeven analysis or market sizing, and then doing full mock cases two or three times per week with video recording so you can catch yourself going silent during calculations. Most people improve noticeably within three weeks of consistent daily practice, but the plateau after that is real and usually means you need to switch to harder numbers or add time pressure, not just do more of the same problems.