Halfway through most cases the interviewer hands you a chart or a table and waits. What they want is not a description of the axes but a reading: the one insight the exhibit contains and what it means for the client. This deck covers the exhibits that come up, from bar charts and line charts to segmented tables and waterfall charts, the traps built into each, from truncated axes to mixed units, and a repeatable way to go from looking at a page to saying something useful about it in thirty seconds.
16 cardsConsultingFree, no signup
Card 1 of 16
What is the interviewer testing when they hand you a chart?
Whether you can extract the one thing that matters and connect it to the case, quickly and out loud.
The chart usually contains one insight the case turns on: one segment is losing money, one competitor is growing, one cost has doubled. The interviewer wants to see whether you find it, how quickly, and whether you can say why it matters for the client's question. They are also watching for discipline: do you read the title and axes before speaking, do you notice the units, do you get lost in the details. A good answer takes half a minute, names the insight, gives one number to support it, and links it to the recommendation. A bad answer describes every bar.
Card 2 of 16
What is the first thing to do with any exhibit, before saying anything?
Read the title, the axes, the units and the source, then take ten seconds of silence to find the pattern.
Say that you would like a moment, then read in order: the title, which tells you what the chart claims to show; the axis labels and units, which tell you what the numbers are; the legend; the time period; and any footnotes, which is where the tricks hide. Only then look at the data, and look for the pattern rather than the points: what is biggest, what is growing, what stands out. Ten seconds of silence is fine and expected. Speaking before you have read the axes is how candidates confidently describe revenue when the chart shows units, or 2019 when it shows 2024.
Card 3 of 16
How do you describe what a chart shows without just listing the bars?
Lead with the headline, support it with one or two numbers, then say what it means; the details can wait for questions.
Structure the reading in three beats. Headline: 'This shows that Ferrow's online sales have tripled since 2021 while store sales have fallen a fifth.' Evidence: 'Online has gone from £100m to £300m; stores from £900m to £720m.' Implication: 'So total revenue is flat, but the mix has shifted sharply toward the channel with lower margins, which may explain the profit decline we are investigating.' That is the whole answer. If the interviewer wants more, they will ask. Listing every data point signals that you cannot tell what matters.
Card 4 of 16
What is a truncated axis, and how does it mislead?
An axis that starts above zero exaggerates differences; check where the axis begins before judging how big a gap is.
If a bar chart's vertical axis starts at 90 rather than 0, a bar at 95 looks twice the height of one at 92, although the real difference is 3%. Charts in company presentations do this constantly, sometimes carelessly and sometimes on purpose. The check takes one glance at the bottom of the axis. In an interview, noticing it and saying 'the axis starts at 90, so the differences are smaller than they look, about 3%' is exactly the kind of observation that earns credit, because it shows you read the chart rather than the picture.
Card 5 of 16
A chart shows revenue growing every year. What questions should you ask before concluding the business is healthy?
Growth in what units, against what, with what margin, and whether it is organic or bought.
Revenue growing does not mean profit growing: ask about margin. It does not mean the company is winning: ask how fast the market grew, since 5% growth in a market growing 15% is losing share. It does not mean the business is doing it: ask whether growth came from acquisitions or currency. And check the units: is it nominal, so inflation is inside it; is it in a currency that moved. A candidate who sees an upward line and says the company is doing well has fallen for the chart. One who asks two of these questions has shown judgement.
Card 6 of 16
What is a waterfall chart, and how do you read one?
It walks from a starting value to an ending value through the steps that changed it; read the biggest step first.
A waterfall starts with one bar, say last year's profit, then shows a floating bar for each factor that added or subtracted, ending with this year's profit. It answers 'what happened between then and now'. Read it by finding the largest positive and largest negative steps, since those are the story: profit fell from £50m to £30m mostly because raw material costs rose £25m, partly offset by £8m of price increases. Ignore the small steps until asked. Waterfalls are how consultants present profit bridges, so being fluent with one shows familiarity with the work.
Card 7 of 16
A table shows profit by customer segment: large accounts £40m, mid-size £25m, small £-15m. What do you say?
Small customers lose money, and the question is why and whether to fix, reprice or exit them.
The headline is that one segment is destroying value: small customers lose £15m, which wipes out more than a third of what the other two earn. Then the questions: is it a cost-to-serve problem, where each small account costs more to handle than it generates; is it a pricing problem, where small customers get the same discounts as large ones; or is it strategic, where small customers grow into large ones and the losses are an investment. The recommendation depends on which. Saying that, rather than just reading the three numbers, is what the exhibit is for.
Card 8 of 16
What is the difference between absolute and relative numbers on a chart, and why does it matter?
A big percentage on a small base can be a small number; always ask what the base is.
A chart showing a segment growing 80% a year looks like the story, until you learn it is £2m of a £2bn business. Conversely, a segment growing 3% might be adding £30m. Charts that show percentages hide the base; charts that show absolute values hide the rate. When you see either, ask for the other. In a case, the common trap is a fast-growing small segment presented as the answer to a big problem. The question to ask aloud is 'what is that as a share of total revenue', and the interviewer will usually tell you.
Card 9 of 16
Two lines on a chart, market size and the client's revenue, both rising. How do you tell whether the client is gaining or losing share?
Compare the growth rates, or compute the ratio at the start and end; a slower-growing line is losing share even while rising.
Rising revenue in a rising market says nothing about share until you compare rates. If the market went from £10bn to £15bn, up 50%, and the client from £1bn to £1.2bn, up 20%, the client's share fell from 10% to 8%. Compute the share at both ends and say it. That is the insight the chart was built to test: a company can be growing and losing at the same time, and a client who is celebrating growth may need to hear that competitors are growing faster. Then ask who is taking the share and why.
Card 10 of 16
What is a stacked bar chart, and what does it tend to hide?
It shows the total and its composition, but makes it hard to read changes in the middle segments.
A stacked bar piles segments on top of each other so the full bar is the total and each band is a part. It is good for seeing the total and the bottom segment, whose base is the axis, and bad for the segments in the middle, whose position depends on everything beneath them. A middle segment can look like it is growing when it is only being pushed up by the one below. If the case turns on a middle segment, ask for the numbers or mentally subtract. Saying 'the middle band is hard to read from a stacked chart, can I have the figures' is a legitimate and impressive request.
Card 11 of 16
An exhibit shows costs per unit falling as volume rises. What is it showing, and what is the trap?
Economies of scale: fixed costs spread over more units; the trap is assuming the curve continues forever.
Cost per unit falls with volume because fixed costs are spread more thinly, and sometimes because buying power and learning improve. That supports a strategy of growing to cut costs. The trap is extrapolating: at some point the factory is full, the next unit needs a new factory, and cost per unit jumps. Diseconomies also appear, as coordination gets harder. So read the chart, name the mechanism, and then ask where the current capacity limit is and what the step cost of the next one would be. That turns a chart into a decision.
Card 12 of 16
The chart is in a currency and time period different from the rest of the case. What do you do?
Say so, convert or align if you can, and if you cannot, state the assumption you are making.
Mixed units are a deliberate test. If the case has been in pounds and the exhibit is in dollars, note it and ask for or estimate an exchange rate before comparing anything. If the exhibit covers 2020 to 2023 and the case question is about 2024, say that you are assuming the trend continued, or ask for 2024. Interviewers reward the noticing more than the conversion. The failure is to compare a dollar figure with a pound figure as if they were the same, which happens surprisingly often under pressure.
Card 13 of 16
How do you go from a chart insight to a recommendation in the same breath?
Insight, cause, action: what the chart shows, why it is probably happening, and what the client should do about it.
After the headline reading, add two sentences. First, a hypothesis for the cause: 'Store sales are falling, most likely because footfall is shifting online rather than because of pricing, given online is growing.' Second, an action: 'That suggests the client should look at closing the weakest stores and investing in fulfilment rather than discounting in-store.' You have then turned an exhibit into a piece of the answer. Even if the hypothesis is wrong, the interviewer sees you reasoning toward a decision, which is the job. Ending on a description leaves the work to them.
Card 14 of 16
What is a scatter plot with a trend line telling you, and what is the classic mistake in reading it?
It shows whether two things move together; the mistake is assuming one causes the other.
A scatter plot puts one variable on each axis and each point is one observation, one store or one customer. A trend line shows correlation: stores with more staff have higher sales, say. The classic error is to conclude that adding staff would raise sales, when it may be that busy stores get more staff. Say what the chart shows, note the correlation, and then ask what might be driving both, or whether there is a natural experiment. Interviewers set this up deliberately, and a candidate who says 'correlation, not necessarily causation, and here is why' stands out.
Card 15 of 16
An exhibit contradicts something the interviewer said earlier in the case. What do you do?
Point it out politely and ask which to rely on; the contradiction may be the point.
Sometimes the interviewer said the market was growing and the exhibit shows it shrinking. That is either a test or a mistake, and either way you should raise it: 'This shows the market declining 2% a year, which is different from the growth we discussed earlier; should I rely on the exhibit?' Never quietly pick one, and never pretend not to notice. Consultants are paid to catch inconsistencies in client data, and an interview is the first place to show that you will. Most interviewers will smile and tell you which is right.
Card 16 of 16
Give a complete thirty-second reading of an exhibit, in order, that would satisfy any interviewer.
Read silently, headline, two numbers, implication, hypothesis, next step.
Take ten seconds and read the title, axes, units and footnotes. Then: 'This chart shows Ferrow's profit by channel from 2021 to 2024. The headline is that store profit has halved, from £80m to £40m, while online has gone from a £10m loss to break-even. Total profit is down £50m, almost all of it from stores. My hypothesis is that fixed store costs have not come down as sales moved online. I would like to test that by looking at store-level sales and costs next.' Headline, evidence, implication, cause, next step. Practise it on any chart in a newspaper until it is automatic.
Reading is not the same as remembering. Study this deck with scheduled review so the cards come back before you forget them.