AND filtering reduces rows; OR expands them. Derived metrics require mental math. Rankings differ by metric. Boundary values flip answers.
Advanced TA questions require filtering on one column and evaluating another — or ranking on a derived value not shown in the table. The strategy: sort once, then scan for the relevant subset.
| Retailer | Revenue ($M) | Stores (#) | Online % | EBITDA ($M) |
|---|---|---|---|---|
| Alpha | $820 | 210 | 38% | $92 |
| Beta | $440 | 95 | 61% | $55 |
| Gamma | $1,100 | 340 | 22% | $88 |
| Delta | $270 | 60 | 75% | -$12 |
| Epsilon | $590 | 145 | 45% | $71 |
Pro Tip: For derived metrics, mentally compute quick estimates. Alpha: $820M / 210 stores ≈ $3.9M/store. Gamma: $1,100M / 340 ≈ $3.2M/store. Alpha has higher revenue per store despite lower total revenue.
A company ranked #1 in total revenue may be ranked #3 in revenue per store and #5 in EBITDA margin. Ranking depends entirely on which metric is used. GMAT often contrasts rankings on different columns to test this.
The company with highest revenue may have the lowest revenue per employee if it also has the most employees.
When both conditions must hold, verify EVERY qualifying row meets both — not just the first one you check.
Don't try to compute exact derived values. Estimate and compare. Precision beyond ±5% is rarely needed.
A company with EBITDA $92M on $820M revenue = 11.2%. One with $71M on $590M = 12.0%. Lower EBITDA, higher margin.
If one cell is blank or N/A, that retailer cannot satisfy conditions involving that column. Don't assume a value.
When sorting EBITDA ascending, negative values appear at the top. Don't misread a negative as a small positive.
At most 3 means ≤3. Fewer than 3 means <3 (i.e., at most 2). One threshold difference can flip the answer.
Online revenue = Revenue × Online%. Don't confuse online % of revenue with online % of customers.
An AND condition with two restrictive filters might yield zero qualifying rows — making an "all of these" statement trivially True (vacuously true) or False depending on context.
If the question asks "more than 3" and you count exactly 3, the answer is False. Boundary values are frequent traps.
Using the retailer table (Alpha: Rev $820M, 210 stores, Online 38%, EBITDA $92M). Which retailer has the highest revenue per store?
Using the same table. Statement: All retailers with online share above 50% have negative EBITDA.
Which retailer has the highest EBITDA margin (EBITDA/Revenue)?
How many retailers have BOTH revenue above $400M AND more than 100 stores?
Statement: The retailer ranked first in total revenue is also ranked first in EBITDA.
Which retailer has the highest online revenue (Revenue × Online%)?
A TA question asks "How many companies have revenue per store greater than $4M?" Based on the table, the answer is:
Statement: Gamma has both the highest revenue and the most stores among the five retailers.
If you need to find "which retailer has the highest EBITDA margin," but EBITDA margin is not a column in the table, you should:
Statement: Exactly two retailers have a negative or zero EBITDA.
This reduces the candidate rows, making the second check faster.
Quick mental division to ±5% accuracy is sufficient for ranking questions.
The top revenue company may rank 3rd in profit margin. They are different questions.
Sorting EBITDA ascending brings losses to the top. Don't misread as small positives.