1. Orient
Title, units, time period, what each axis or column is.
- Ask
- "Millions or thousands?"
To read a case interview exhibit, first check the title, units and time period, then find the one or two biggest patterns, calculate only what proves them, and say what they mean for the client's question. In our practice cases the so what and the next step carry the score, not how many numbers you read aloud. McKinsey's practice cases tell you to look for trends that are not immediately obvious from the data. Soreno, an AI voice interview practice app for consulting, investment banking and other finance candidates, runs free case math drills with no sign-up, for the growth rates and shares exhibits ask for, and full mock cases with scored feedback.
Updated
Tamarack Outdoor, an invented outdoor gear maker, grew fast last year, but the CEO says profit is down. The interviewer hands you one exhibit: revenue and operating margin by product line for 2024 and 2025. What does it tell you? Every number below is invented by Soreno for practice. Try your own read before the steps that follow.
| Product line | Revenue 2024 | Revenue 2025 | Margin 2024 | Margin 2025 |
|---|---|---|---|---|
| Tents $120M
| $120M | $110M | 25% | 25% |
| Jackets $80M
| $80M | $84M | 20% | 19% |
| Footwear $50M
| $50M | $106M | 4% | 2% |
| Total $250M
| $250M | $300M | 19.2% | 15.2% |
Revenue in $ millions. Margin is operating profit as a share of revenue. Source: invented case data.
No firm publishes a method for this, so these six steps are ours. The first two take seconds. In our practice cases most of the credit sits in steps four to six, which is where most candidates stop short.
Title, units, time period, what each axis or column is.
The one biggest thing the exhibit shows.
Which category, year or segment breaks the pattern.
Only the numbers that prove or kill the headline.
What it means for the client's question.
What you would ask for or look at next.
The table above, read the way we would read it in a case. Every number is computed from it.
1. Orient
Revenue is in $ millions and margin is operating profit over revenue, so profit per line is revenue times margin. The exhibit has no costs and no unit volumes, so anything about price or units is a question for later, not a reading of this table.
2. Headline
Total revenue went from $250M to $300M, up 20%. Profit went from $48 million to $45.58 million, down 5%, and the overall margin fell from 19.2% to 15.2%. Growth without profit is the headline.
3. Compare
Footwear more than doubled, from $50M to $106M, while its margin fell from 4% to 2%. Tents, the most profitable line at 25%, shrank from $120M to $110M. Footwear went from 20% to 35.3% of revenue: a mix shift toward the thinnest margin.
4. Calculate
Tents: $30 million to $27.5 million, minus $2.5 million. Jackets: $16 million to $15.96 million, minus $40,000. Footwear: $2 million to $2.12 million, plus $120,000. Together minus $2.42 million. Footwear's profit rose only $120,000 on $56M of extra revenue, about 0.2 cents a dollar, because its margin also fell on the sales it already had.
5. So what
It is tempting to blame footwear. But footwear's profit barely moved. The drop is the tents: their $2.5 million decline is more than the whole $2.42 million fall. The real problem is twofold: the best line is shrinking, and the fast growth in footwear adds revenue with almost no profit behind it.
6. Next step
Why did tent sales fall (price, a competitor, or marketing moved to footwear)? And how much of footwear's growth was bought with discounts? Those two answers decide whether to defend tents, reprice footwear or both.
Answer first, then the numbers that prove it, then the next step.
Say it
"The headline is growth without profit: revenue grew 20% but profit fell about 5%. Two things drive it. First, tents, their best margin at 25%, shrank by $10M, which alone costs $2.5 million, more than the whole drop. Second, footwear more than doubled but earns only 2%, so its $56M of new revenue added almost nothing. Next I'd want to know why tent sales fell, and how much of footwear's growth came from discounts."
Weak version
"Tents were 120 and now 110, jackets were 80 and now 84, footwear was 50 and now 106, and the margins were 25, 20 and 4 and now 25, 19 and 2." Every number is right. None of them has been turned into a finding.
No firm publishes a list of the charts it uses. These are the ones we see most in practice cases, with the trap in each.
Bar
Check whether the axis starts at zero. A bar chart that starts at 90 makes a 5% gap look like a doubling.
Line
Check the time step and whether two lines share one axis. A second axis on the right can make any two lines look linked.
Stacked bar
Only the bottom segment sits on a common baseline. Compute the middle segments rather than eyeballing them.
100% stacked or pie
Shares can rise while the amount falls. Ask for the total before you call a segment growing.
Waterfall
Read it as a bridge from start to end. The biggest bar is usually the answer to "why did it change?".
Mekko
Width is segment size, height is share. Multiply the two to get each player's revenue in a segment.
Table
Scan for the row or column that breaks the pattern, then compute only that. The Tamarack exhibit above is a table.
Both come up often in our practice cases.
Small base
Footwear grew 112% and tents fell only 8.3%. Judged by percent, footwear is the story. Judged by profit, tents moved $2.5 million and footwear $120,000. Always convert to dollars before you rank.
Averages
The total margin fell from 19.2% to 15.2%, yet tents held at 25% and jackets fell only one point. At 2024 margins, this year's mix alone would give 16.2%, so about 75% of the fall in the average is mix: more revenue in the low margin line. A falling average does not mean every line got worse.
Growth rates, shares, margins and breakevens, timed, no sign-up. The calculations above are the kind these drills train.
Do the math in your head or on paper, type the answer and press Enter. You see right or wrong straight away, then your score and average time per question.
Two questions from each type: percentages, growth rates, break-even, margins and large numbers. Answers within 1% count as correct.
Four things we score in practice cases. Reading numbers aloud earns none of them.
Say the finding before the figures. "Growth without profit" beats a tour of every cell.
Two or three calculations that prove the headline. Leave the rest of the table alone.
Tie the finding to the case question and your current hypothesis. That is what turns an observation into an insight.
Name the data you would ask for next and why. It shows the read is moving the case forward.
Each one usually comes with a chart or table mid-case. Under each is what a strong read does.
Revenue and cost exhibits.
An airline is flying more passengers but making less money. Why?
Reads the revenue and cost per passenger exhibit for the line that moved, then computes it per seat before naming the cause. Practice it, see a model answer
A software company is growing fast but losing money. Should it worry?
Uses the cohort or cost table to split growth spending from losses on each customer, then answers yes or no. Practice it, see a model answer
A grocery chain's profits fell 20% in a year. What happened?
Turns the store and category exhibit into a profit change by line, the way the Tamarack read does. Practice it, see a model answer
From practice cases. McKinsey's hints point at the second.
Thousands read as millions, or a percent read as a percentage point, wrecks every number that follows. Say the units out loud first.
The first pattern you see is often not the answer. In Tamarack, footwear jumps out, but tents explain the profit drop.
Working out every growth rate burns minutes. Decide what you are testing, then compute only that.
Take a moment, but tell the interviewer what you are looking for. A short narration lets them help you.
An exhibit read that never returns to the client's question leaves the interviewer to finish it for you.
None of these pages gives a chart-reading method or a time limit. They do say what the read should show. Read on October 31, 2026.
McKinsey's practice cases put data in front of you and ask what it tells you. The hints say to take some time to look at the information, note down observations and challenge yourself to identify trends that are not immediately obvious. They also say calculators are not allowed and you need not feel rushed into calculations.
BCG says you will analyze data, perform quick calculations and identify the most important factors, and that what matters most is how you approach the problem and the quality of your reasoning.
Oliver Wyman asks whether you can evaluate data and use it in your analysis, and tells candidates to analyze and incorporate data into their thought structure and to be careful with simple calculations.
L.E.K. tells candidates to share their thought process as information is revealed, and to commit to a course of action while showing a willingness to rethink in the face of new data.
Soreno's AI interviewer runs full cases with charts and tables, asks for your so what after each one and scores the session. Start your free week.
A chart, table or slide of data the interviewer hands you mid-case. You are expected to read it, find what matters and say what it means for the client's question.
No firm publishes a limit. McKinsey's hints say to take some time to look at the information. In our experience a quiet half minute to a minute, then a read of about a minute, works, with a short narration if you need longer.
No. Calculate only what proves or kills your headline, usually a growth rate, a share or a ratio. McKinsey's practice cases say calculators are not allowed, so keep the math simple and write it down.
Go back to the question the exhibit was meant to answer, then compare the biggest category to the rest. If nothing stands out, say what you'd expect to see and ask what the interviewer wants you to focus on.
No firm publishes a list. In our practice cases bar, line, stacked bar, waterfall, Mekko and plain tables come up most. Each has its own trap, listed above.
An observation says what the data shows: footwear revenue doubled. An insight says what it means for the client: growth is coming from the line that earns almost nothing.
Yes. Clarifying units, definitions or what's missing is part of a good read. L.E.K. tells candidates to clarify issues and ask for the information they need.