Start with a decision that can go wrong
A team proposes a 10% price cut and expects unit sales to rise 20%. Revenue would grow, so the slide calls the promotion a success. A pricing intern should be able to ask the next question: what happens to contribution after the variable cost of those extra units? That question makes a useful work sample because the arithmetic is simple but the commercial interpretation is not.
BLS describes market-research work as investigating customers, market conditions and factors affecting demand, including prices. Pricing internships may sit in a more specialized team, but that background helps explain the connection between customer evidence and commercial analysis. Use the actual posting to determine whether the assignment concerns retail promotions, subscription packaging, contracts, competitive research or another pricing problem.
Search pricing analyst intern, pricing strategy intern, revenue management intern, commercial analytics intern and promotion analytics intern. Those titles are not interchangeable. An airline revenue-management role may involve capacity and booking timing; a retailer may focus on item prices and promotions; a software company may focus on plans and contracts. Choose the setting before designing your project.
Sources[1]
Define price, cost and the outcome you are optimizing
For the fictional case, an item sells for $40, has $22 of variable cost per unit and sells 1,000 units during a defined period. Revenue is $40,000 and contribution before fixed costs is ($40 − $22) × 1,000 = $18,000. The proposed price is $36, with an assumed 1,200 units sold and the same variable cost per unit.
Be explicit about what the $22 includes. In a real analysis, shipping, payment fees, returns, fulfillment labor and commissions may vary differently with sales. A simplified exercise can hold these costs constant, but it must not claim to model every expense. Contribution before fixed costs is not net income, and a promotion's incremental contribution is not the same as total company profit.
Also define the horizon. A one-week promotion can move purchases forward from the following week rather than create additional demand. A discount may acquire new customers, clear inventory or support another objective. You cannot evaluate those goals with one formula unless you also model their relevant benefits and costs. State the narrow question your first calculation answers and add other effects separately.
Reconcile the revenue gain and contribution loss
At the proposed price and volume, revenue is $36 × 1,200 = $43,200, an increase of $3,200 or 8%. Contribution is ($36 − $22) × 1,200 = $16,800, a decrease of $1,200 or about 6.7%. Higher revenue and more units coexist with lower contribution because the contribution per unit falls from $18 to $14.
To preserve the original $18,000 contribution at $14 per unit, the team would need approximately 1,285.7 units. If only whole units can be sold, it needs at least 1,286. Relative to 1,000 units, that is an increase of about 28.6%, not 20%. Do not round the break-even quantity down to 1,285 and call it sufficient; 1,285 × $14 is $17,990.
This calculation does not prove that the discount should be rejected. It shows the sales response required under the stated assumptions. A different variable cost, a vendor-funded promotion, avoided inventory write-offs or follow-on customer contribution could change the decision. Put those items in separate lines with evidence. Do not add vague strategic value until the proposal appears attractive.
Understand elasticity without pretending you estimated demand
OpenStax explains price elasticity using percentage changes and a midpoint method that uses average starting and ending values. For our two fictional price-volume points, the midpoint quantity change is 200/1,100 and the midpoint price change is −4/38. Their ratio is about −1.73. The calculation is a description of the supplied points, not a validated causal estimate of how buyers would respond.
In this case the 1,200-unit volume was an assumption. Calling the resulting ratio measured elasticity would hide that fact. Even with two observed periods, a seasonal event, marketing campaign, stockout or competitor change could affect sales. Explain whether the analysis uses a demand model, a controlled test, a natural experiment or simply a scenario. Each supports a different strength of conclusion.
For a first portfolio, it is often better to present a transparent sensitivity table than a fragile regression with an impressive coefficient. Calculate contribution at 1,100, 1,200, 1,300 and 1,400 units under the proposed price. At $14 contribution per unit, those results are $15,400, $16,800, $18,200 and $19,600. The table makes the decision threshold visible without pretending to know which outcome will occur.
Sources[2]
Add the practical constraints that change a promotion
Suppose only 1,250 units can be supplied during the promotion. Under the simplified model, the proposed price cannot preserve the original contribution because maximum contribution would be $17,500. That is a useful constraint to surface before debating a more optimistic demand estimate. A demand forecast above available stock is not a feasible sales plan unless replenishment is included.
Next consider whether the promoted product substitutes for a higher-margin item. Additional units of the discounted product may not be additional units for the company. Likewise, an advertised price may involve eligibility rules, bundles or channel-specific fees. Keep the price actually paid and the unit of sale consistent across comparisons. A competitor's subscription price, for example, may include services absent from your product.
For real employer work, do not independently change live prices or promise promotional terms. A portfolio should show analysis and a proposed decision, not an unauthorized commercial action. Use synthetic data or properly licensed public datasets, and state what customer or transaction information would be needed to extend the work.
Write a recommendation that survives a follow-up question
A useful memo might read: the proposed price cut increases modeled revenue 8% but reduces contribution about 6.7% at the assumed volume. It requires at least 1,286 units to preserve contribution, before any added promotion cost. The next decision is whether credible demand and supply evidence supports that threshold. That is more helpful than either discounts are bad or revenue goes up, so proceed.
Include the strongest counterargument to your own recommendation. Perhaps inventory will expire or the vendor reimburses part of the discount. Show how the answer changes when that fact is added. This demonstrates commercial judgment rather than a mechanical attachment to one metric. Keep a clear distinction between an observed fact and the scenario you introduced to explore it.
In an interview, expect a reviewer to change a number. Be ready to recompute the threshold when variable cost becomes $20 or the price becomes $38. The point is not mental-arithmetic theater. It is whether you understand the relationship well enough to adapt the model without rebuilding it from scratch or confusing contribution with revenue.
Match the project to the internship
Read for the actual mix of analytical and commercial work. A role may require spreadsheet modeling, SQL, competitor research, presentation skills or statistical coursework. Do not assume that every pricing internship requires advanced econometrics, and do not ignore a technical requirement stated in a particular posting. Show the relevant part of your project first.
Ask how the team evaluates a pricing recommendation, what data interns can access, and who reviews assumptions. Clarify whether the work concerns list prices, realized prices, discounts, bundles or capacity allocation. These distinctions determine which project would help you prepare and which questions are worth asking before accepting.
Your finished sample should contain a small input table, checked formulas, a sensitivity table and a short decision note. It should be possible to reproduce the result in a few minutes. That is a stronger application asset than a broad pricing-strategy deck whose central recommendation depends on an unexplained volume assumption.