Introduction
“Revenue grew by 20% last month, but operating profit fell by 40%. What would you investigate?”
A useful answer starts by defining the numbers and showing how they connect. More customers can increase revenue while lower realised prices or higher service costs reduce what each customer contributes.
This data analytics case interview walks through a fictional subscription business. You will clarify the problem, request the right data, reconcile the profit movement and explain a recommendation. All figures are illustrative, in USD. Try answering each stage before reading the calculation.
Clarify What the Interviewer Means by Profit
Start with three questions:
- Are we discussing gross profit, operating profit or net profit?
- Are the periods comparable in scope, currency and accounting treatment?
- Does management want to explain the decline, restore profit or assess whether growth is worth its cost?
For this case, the interviewer confirms operating profit and one subscription plan. Revenue relates to service delivered during each month, after discounts and refunds. Variable and fixed operating costs cover all relevant expenses once, including applicable depreciation. Interest and income taxes are excluded.
The volume measure is paid account-months, meaning full-month equivalents of paid service. It is not simply the number of accounts active on the last day of the month. Revenue and service costs must relate to the same activity.
This definition prevents a common subscription-analysis mistake: comparing a full month’s revenue with an end-of-month customer count.
Ask for the Data That Can Explain the Change
Request paid account-months, net revenue, variable service costs and fixed operating expenses for both periods. Then ask for billing details, discounts, refunds, usage and customer cohorts if the totals reveal a change worth investigating.
The data checks behind SQL business problems apply here: aggregate billing and cost records to the same account-month before joining them. Joining several invoice lines to several usage records can multiply both totals.
Ask whether shared expenses are allocated consistently and reconcile the result to finance’s operating statement. For this exercise, the interviewer supplies reconciled figures:
| Metric | Month 1 | Month 2 |
| Paid account-months | 1,000 | 1,500 |
| Net revenue per account-month | $100 | $80 |
| Variable cost per account-month | $40 | $48 |
| Fixed operating costs | $30,000 | $30,000 |
| Total net revenue | $100,000 | $120,000 |
| Operating profit | $30,000 | $18,000 |
| Operating margin | 30% | 15% |
Volume grew by 50%, while net revenue per account-month fell by 20%. Those movements together explain the 20% revenue increase.
Build the Worked Profit Breakdown
Contribution is revenue remaining after variable costs. It covers fixed operating costs, with the remainder becoming operating profit in this model.
Month 1 contribution is 1,000 multiplied by $60, because each account-month earns $100 and incurs $40 of variable cost. Deducting $30,000 of fixed costs leaves $30,000 operating profit.
Month 2 contribution is 1,500 multiplied by $32, because $80 minus $48 leaves $32 per account-month. Deducting the same fixed costs leaves $18,000 operating profit.
That establishes the $12,000 decline. Next, explain the movements behind it.
Measure the Volume Effect
Hold Month 1’s unit economics constant. The additional 500 account-months would contribute $60 each.
Volume effect: 500 × $60 = +$30,000.
This is a comparison at the earlier contribution rate. It does not mean the extra accounts actually delivered that amount under Month 2’s prices and costs.
Measure the Realised Price Effect
Now apply the change in net revenue per account-month to Month 2’s volume. Here, “realised price” means the average revenue retained per unit of service, not the advertised price.
Realised price effect: 1,500 × ($80 − $100) = −$30,000.
This offsets the favourable volume effect in the breakdown.
Measure the Variable Cost Effect
Variable cost increased by $8 per account-month. Apply that change to the current volume.
Variable cost effect: −1,500 × ($48 − $40) = −$12,000.
Fixed operating costs did not change. The complete reconciliation is:
$30,000 starting profit + $30,000 volume − $30,000 realised price − $12,000 variable cost = $18,000 ending profit.
The order matters: volume is valued at earlier unit economics, then price and cost changes are applied to current volume. State that convention and use it consistently. Other conventions can distribute interaction effects differently, but the total change must reconcile.
Operating margin fell from 30% to 15%, a decline of 15 percentage points. Operating profit fell by 40%. The business remains profitable.
Separate Findings From Possible Causes
The figures establish that contribution per account-month fell from $60 to $32. They do not establish why.
Lower realised revenue could reflect deeper discounts, refunds or a larger share of customers on promotional terms. Check invoice-level changes and compare customers with similar tenure and contract conditions.
Higher variable costs could reflect supplier price increases, heavier usage or temporary onboarding work. Compare usage per account with cost per unit of usage before calling the change inefficiency.
Investigate the realised-revenue change first because its adverse effect is larger in this breakdown, while examining service costs alongside it. Assess newer customer cohorts over an appropriate period before deciding whether acquisition or onboarding costs are worthwhile.
Recommend an Action and Explain Its Assumptions
Suppose billing analysis identifies a discount worth testing. Model a narrower offer that raises net revenue per account-month from $80 to $88.
At 1,500 account-months, a $48 variable cost rate and unchanged fixed costs, contribution would be $60,000 and operating profit would return to $30,000.
That is a scenario, not a forecast. If volume fell by 10% to 1,350 account-months, profit would instead be $24,000, assuming the same cost rates. It would improve on the current $18,000, but would not fully restore the earlier result.
When designing an A/B test, define eligible accounts, assignment and the measurement window in advance. Compare contribution per eligible account, including those who do not convert or renew. Track conversion, retention, refunds and service quality alongside it. Confirm whether implementation adds costs before estimating total operating profit.
Give a Concise Interview Answer
Here is a sample response after receiving the data:
“Revenue grew because paid account-months increased from 1,000 to 1,500, even though net revenue per account-month fell from $100 to $80. Variable cost per account-month also rose from $40 to $48, reducing unit contribution from $60 to $32.
“Using earlier unit economics for volume, growth adds $30,000 of contribution. Lower realised revenue removes $30,000, and the higher variable cost rate removes another $12,000. Fixed costs are unchanged, so operating profit reconciles from $30,000 to $18,000.
“I would first investigate discount and refund changes, then separate usage growth from supplier-rate changes. The totals identify financial effects, but do not prove the underlying causes.
“If the evidence supports narrowing a discount, I would test it on eligible accounts and measure contribution alongside retention. An $88 realised rate restores earlier profit only if volume and costs hold. With a 10% volume decline, profit reaches $24,000 instead. I would present both outcomes and validate the assumptions before recommending rollout.”
Avoid saying simply that “costs rose faster than revenue”. That describes the result without explaining the drivers. Also avoid treating a margin decline as an operating loss or presenting the pricing scenario as a guaranteed recovery.
Practise Explaining the Decision
Before sharing your answer, check that you have:
- Defined the metric, reporting period and unit of analysis.
- Reconciled the full profit movement without double counting.
- Separated observed changes from unverified explanations.
- Connected the recommendation to assumptions and measurable outcomes.
CompeteX is PangaeaX’s platform for practical data competitions. Its data analytics challenges offer a way to practise defining metrics, validating data and explaining findings across unfamiliar business problems.
Use this case as a starting point. Change one input and recalculate the answer, then explain aloud what changed in your recommendation. The reasoning should remain clear even when the numbers are unfamiliar.