Most candidates see an exhibit and start reading every number out loud. Thirty seconds later they have narrated the chart and still have not said what it means. Interviewers are not scoring whether you can read. They are scoring whether you can find the one thing on the page that changes the call.

Read the frame first, find the one gap, say the so-what, then ask one next question. That is the whole job, and it fits in about 45 seconds.

Why exhibits trip people up

• They read before they orient. Without the title, units, and time period, a number can mean the opposite of what you think.
• They treat every number as equal. Most exhibits have one row, bar, or trend that matters and a lot of context around it.
• They stop at the observation. "Segment B fell 12 percent" is a fact. The interviewer is waiting for what it means for the recommendation.

The 45-second exhibit sequence

1) Frame it in 10 seconds

Say the title, the units, and the time period out loud before you touch a number. "This is revenue by region, in millions, over the last three years." That one sentence prevents most misreads, like treating a percentage as a dollar figure or a share as a total.

2) Find the one gap in 15 seconds

Scan for the single biggest outlier: the bar that moved most, the row that breaks the pattern, or the trend line that bends. Ignore the rows that behave like everything else. If you need a quick number, calculate only the one ratio or difference that sizes the gap, and round hard.

3) Say the so-what in 10 seconds

Tie the gap back to your working hypothesis. Fact, meaning, implication: "The Northeast drove almost all of the decline, so this is a regional problem, not a company-wide pricing issue."

4) Ask one next question in 10 seconds

Name the single piece of data that would confirm or kill what you just said. "Do we know whether the Northeast lost customers or lost spend per customer?" One question, not a list.

Want reps on this? Practice exhibit reading with CaseCrunchAI and get feedback on whether you found the gap and said the so-what before time ran out.

A 45-second script

Exhibit: a table of unit sales and average price for four product lines over two years.

"This shows unit sales and average price for four product lines, last year versus this year. Three lines are roughly flat. Line C is the outlier: units are up about 30 percent but average price is down about 20 percent, so revenue on C is only up a few percent while it is eating most of our volume growth. That tells me the growth is discount-driven, which supports the margin hypothesis. Next, I would want to see the margin by line to check whether C is now below our cost to serve."

Frame, gap, so-what, next question. Nothing else.

What to skip

• Reading the table row by row. If you are narrating, you are not analyzing.
• Calculating everything. Calculate the one number that sizes the gap. Leave the rest.
• A so-what that does not touch the recommendation. "Interesting trend" is not a so-what. Say what it changes.
• Five follow-up questions. Ask the one that would most change your answer.

10-minute drill

1. Grab three charts or tables from old cases or business news.
2. For each, set a 45-second timer. Out loud: frame it, name the one gap, say the so-what, ask one next question.
3. Replay or rewrite what you said. Cut any sentence that just restated a number without meaning.
4. Redo the exhibit where you ran out of time before the so-what.

An exhibit is not a reading test. Find the one gap, say what it means, and ask for the data that would prove it.

Start practicing free at CaseCrunchAI and run timed exhibit drills that force a so-what before the clock runs out.