Spotting the Trap in an Exhibit
Two candidates stare at the same exhibit: revenue is rising, profit is falling. One says, "Costs must be increasing"; the other pauses, checks the denominator, notices the customer mix has shifted, and finds the real issue in thirty seconds.
That pause is the skill. Exhibit traps are not about hard math - they are about resisting the first obvious story a chart wants you to tell.
- An exhibit trap is a plausible but wrong conclusion caused by scale, mix, base, denominator, timing or missing context.
- Never describe a chart first. First ask: What is being measured, against what base, over what time, and for which segment?
- The five classic traps are: wrong denominator, mix shift, index/base effect, correlation-causation, and missing benchmark.
- If totals improve but unit metrics worsen, suspect volume growth hiding poor economics.
- If averages look stable, check segment-level data - weighted averages can hide the real movement.
- Your answer should separate: observation - implication - hypothesis - next data needed.
- The safest line in an exhibit case is: "Before concluding, I would normalize this by the relevant base."
Big Picture - The Exhibit Is Not the Answer
A case exhibit is evidence, not truth. Your job is not to read it aloud; your job is to test whether the visual supports the business conclusion being suggested.
Think of every exhibit as having two layers: the visible layer shows numbers, bars, lines or tables; the hidden layer contains definitions, denominators, segment mix, timing, and business context. Most traps live in the hidden layer.
Core Explanation - The Five Traps You Must Catch
The first discipline is to avoid reacting to the biggest visual movement. A spike, decline or gap may be real, but it may not mean what it appears to mean. Start with the exhibit title, then the axis labels, then the unit, then the base.
1. Wrong Denominator Trap
This is the most common exhibit trap. The numerator looks impressive, but the relevant denominator tells a different story.
Example: total app downloads rise, but monthly active users as a percentage of downloads fall. The business problem is not acquisition; it may be activation, retention or product-market fit.
2. Mix Shift Trap
A total or average changes because the composition changed, not because performance improved or worsened in each segment.
Example: average margin falls even though each product line has stable margins, because the company sold more of the low-margin product. In profitability cases, mix shift is often more important than price or cost alone.
3. Index or Base Effect Trap
An indexed chart makes growth look dramatic or weak depending on the chosen base year. A small business growing from a tiny base may show a steep curve but still contribute little absolute value.
When you see an index, ask: "What are the actual values behind the index?"
4. Correlation-Causation Trap
Two lines moving together do not prove one caused the other. Advertising spend and sales may rise together because both respond to seasonality, not because the campaign worked.
Your safe wording: "This suggests an association, but I would need a control, time lag or comparable market to infer causality."
5. Missing Benchmark Trap
A metric can look good in isolation and weak against the right benchmark. A 12 percent margin may be strong in grocery retail and weak in enterprise software. Without industry, competitor, historical or customer benchmark, the number floats.
This is why defining the problem before solving it matters: the benchmark depends on the decision you are trying to make.
Definitions You Can Say in One Breath
- Exhibit: A case document, chart, table or visual that provides evidence for solving a business problem.
- Exhibit trap: A plausible wrong conclusion caused by hidden base, mix, timing, benchmark, causality or measurement issues.
- Denominator: The base against which a number is divided to create a rate, ratio or percentage.
- Mix shift: A change in overall performance caused by a change in segment weights, not segment performance.
- Normalization: Converting numbers to a comparable basis so differences reflect real performance, not scale.
The Consultant's Exhibit Checklist
Before you interpret any exhibit, run this checklist silently. It takes less than ten seconds with practice.
Worked Example - When the Average Lies
Suppose a retailer has two formats. Both formats maintain the same margin, but the overall margin falls. Many candidates incorrectly say, "Costs increased." The real answer is mix shift.
The margin did not fall because either format became worse. It fell because the company sold more through the lower-margin format. In a case, that changes your recommendation: you would not start with cost cutting; you would investigate format strategy, pricing, basket size and whether the lower-margin format creates scale benefits.
If the next exhibit asks for break-even impact, revise contribution margin and break-even analysis in cases before attempting the calculation.
Exhibit Sanity Metrics - What to Calculate Fast
These are not universal "good" or "bad" numbers. In a case, a strong number is one that is directionally better than the relevant benchmark: prior period, competitor, target, segment or unit economics threshold.
The trick is not memorising benchmarks. The trick is asking, "Compared to what?" A 5 percent conversion rate can be excellent in one funnel and poor in another.
Case Study - Titan: The Revenue Growth Trap in Jewellery
Titan's jewellery business shows why a revenue exhibit can mislead if you do not separate price, volume, mix and store expansion.
Imagine a case exhibit showing strong revenue growth for a jewellery retailer. The tempting conclusion is simple: "Demand is growing, so expand stores." But in jewellery, revenue can rise for several different reasons: higher gold prices, more wedding purchases, more studded jewellery, more stores, better conversion, higher ticket size, or genuine volume growth.
Titan is a useful Indian example because jewellery performance is not just a pure volume story. The primary driver of growth may be consumer trust and branded retail strength, supported by store expansion, product mix, wedding demand, design freshness, and operating discipline. A revenue-only exhibit hides these drivers.
The move a good candidate makes is to split revenue into components before recommending expansion: number of stores, footfall per store, conversion rate, average transaction value, product mix and gross margin. If revenue is rising mainly because gold prices increased, opening many new stores may be a weak conclusion. If revenue is rising because conversion and ticket size are improving across cohorts of stores, expansion is more defensible.
The lesson: do not recommend from the headline. Decompose the metric until the business action becomes obvious.
How AI Changes Spotting the Trap in an Exhibit
AI makes exhibit practice faster, but it also creates new risks. By 2026, candidates will increasingly face AI-generated charts, dashboards and market summaries in prep - and some will look polished while still having weak logic.
- AI can generate more realistic exhibit drills. You can ask ChatGPT or Claude to create tables with hidden mix shift, denominator or benchmark traps. This helps you practise pattern recognition, not just arithmetic.
- AI can summarize exhibits, but it may miss business context. An LLM may correctly describe a trend yet fail to ask whether the denominator changed. Treat AI as a first reader, not the final consultant.
- AI raises the bar on data skepticism. Dashboards can now be built quickly, so interviewers may test whether you challenge definitions, source quality and metric construction.
Use NotebookLM: upload this lesson, one annual report, and your case notes. Ask it to generate five exhibit questions where the first obvious answer is wrong, then practise answering each in the format: observation, trap, implication, next data needed.
If you want to use AI for live practice rather than passive revision, the natural next step is practising cases with AI as a mock interviewer.
Interview Relevance
"Here is an exhibit showing that our client's online sales have grown faster than offline sales, but profitability has declined. What do you make of this?"
Do not answer with a conclusion immediately. Use a structured exhibit response that proves you can read, interpret and challenge data.
Use this sentence: "The visible trend is X, but before concluding Y, I would normalize it by Z." It signals maturity instantly.
Common Mistake
The mistake: narrating the exhibit instead of challenging it. Candidates say, "Sales are up, costs are up, margins are down," but never ask whether the metric is per customer, per order, per store or total. The fix: after every observation, add one trap-checking question: "What denominator or segment could change this conclusion?"