Pricing Strategy Cases 2026: Recomputed From Two Firms

Three bases in one pricing case, and the spread between them

Oliver Wyman publishes a pricing case at oliverwyman.com, checked 30 September 2026. An Indian startup has built a digital therapy for autistic children and the chief executive wants a price before launch.

The firm names the three bases and puts them in order: "There are three broad ways of pricing the product: value-based, competitor-based, and cost-based." Then it insists on the sequence: "You must first break up pricing strategies into these three categories before moving on with calculations."

Two of those bases are computed on the page. Value first. Trained autism therapy doctors charge around 300,000 rupees a year and the product removes a third of that need, so a third of 300,000 is 100,000 rupees. The firm calls that "the upper limit".

Cost next. Annual fixed costs of delivering the product are about 10 billion rupees. Divide by a user base of 1 million and you get 10,000 rupees per user per year. Add a 50 percent markup and the page reaches "INR 15,000 (INR 10,000 * 1.5)".

So the published range is 15,000 to 100,000 rupees. Divide the top by the bottom: 6.67. One case, one product, and a defensible price band spanning a factor of nearly seven. That spread is the reason the base has to be named before the number is spoken, and it is also why a candidate who opens with cost-plus has quietly given away 85 percent of the available price.

Where the pricing case gets its addressable number, and where it stops

The 1 million users in that division deserve a second look, because it is the one figure a candidate cannot reproduce from the page.

The firm writes: "a realistic goal is to capture 25% of the calculated user base (25% of 4.2 million, which is roughly 1million)". Twenty five percent of 4.2 million is 1.05 million, so the rounding to 1 million is the firm's, and it is fair.

The question is where 4.2 million came from. The page names the filters, income level, internet and tablet penetration, and English-speaking population, and publishes no intermediate number. What you can reconstruct is the first filter only: 10 million autistic children, of whom the middle and high income groups are 40 and 10 percent, so 50 percent, which is 5 million. Getting from 5 million to 4.2 million implies a residual factor of 84 percent covering the remaining filters, and the page never states it.

Two things follow. The 4.2 million is Oliver Wyman's figure and we cite it as such rather than rebuilding it. And the rounding has a price: with the exact 1.05 million users, fixed cost per user is 10 billion divided by 1,050,000, which is 9,524 rupees, and the cost-based price becomes 9,524 times 1.5, so 14,286 rupees. The firm's rounding raises the cost floor by 5 percent.

That is worth saying out loud in an interview, in one sentence, because it demonstrates the thing the firm lists among its own evaluation criteria: "Ability to identify and ask for key information needed, for example the size of the addressable market".

The subscriber count hidden in a penetration price

Now the calculation the page does not run, and it is the one that decides whether the recommendation is safe.

The firm's answer to its second question recommends a subscription "priced between INR 15,000 and INR 50,000 annually", and for reach among lower income families it suggests "between INR 15,000 and INR 30,000". Fixed costs are 10 billion rupees a year and variable costs are stated to be negligible. So the number of subscribers needed to cover fixed cost is fixed cost divided by price.

Read the first row. At the bottom of the recommended penetration band, the business needs 666,667 subscribers before it earns a rupee, which is 63.5 percent of the user base the case itself treats as a realistic capture. Miss the target by a third and the price recommended for affordability stops covering its own fixed cost.

Read the second row against the first. Doubling the price to 30,000 rupees halves the required base to 333,333, under a third of target. That is the whole trade in one line: the affordability argument is real and it costs 333,334 subscribers of safety margin.

This is what a pricing answer should end on. Our page on break-even analysis works the same algebra in the general case, and the breakeven calculator will do the division while you argue about the numerator.

A loyalty program against a discount, in a published pricing case

Bain publishes a different kind of pricing decision in FashionCo at bain.com, checked 30 September 2026. Revenue has fallen for five years and two options reach the chief executive. The question is explicit: "Which one would result in more revenue this year?"

The printed assumptions. Ten million customers, flat. Average pre-sale spend of 100 dollars per person per year, spread evenly across the year. Option A is a rewards program with 25 percent joining in year one, a one-time activation fee of 50 dollars, and a permanent 20 percent discount for members. Option B runs a 20 percent discount for three months a year, and assumes the discount doubles pre-sale purchases over those three months.

Baseline first: 10 million times 100 dollars is 1,000 million dollars.

Option A. Members are 2.5 million. They spend 100 dollars at 20 percent off, so 80 dollars each, giving 200 million. Activation fees add 2.5 million times 50 dollars, so 125 million. Non-members are 7.5 million at 100 dollars, so 750 million. Total 1,075 million dollars.

Option B. Spend is even across the year, so three months carry a quarter of it, 25 dollars a head and 250 million in total at full price. Doubling takes that to 500 million of pre-sale value, and the 20 percent discount collects 400 million of it. The other nine months carry 750 million. Total 1,150 million dollars.

The gap is 1,150 minus 1,075, so 75 million dollars, and Option B wins on revenue. The published recommendation says Option B "is projected to generate $50M more than the awards program (Option A)". Both calculation tables on that page are images we could not read. We looked for a reading of the printed assumptions that produces exactly 50 million and did not find one: treating the activation fee as a cost instead of a receipt gives 200 million, and leaving member spend at 100 dollars net gives 25 million. The direction holds in every reading. Following CONTENT.md section 6.11, both figures stand side by side here rather than one being chosen for the firm.

The pivot that reverses a pricing case answer

The question asked about revenue, and answering it correctly is the right move. The interesting number sits one line below, and it is available from the same assumptions plus one declared rate.

Assume cost of goods runs at 50 percent of pre-sale value. That figure is ours. Pre-sale volume is what drives cost, and the two options move it differently.

Option A adds 75 million of contribution and Option B adds 25 million, so on contribution the ranking reverses. The discount buys revenue by selling 250 million dollars of extra goods at a 20 percent markdown, and the rewards program buys revenue mostly through a fee that carries no goods at all.

Now solve for the rate at which they tie, which is the number worth taking into an interview. Contribution under A is 1,075 minus 1,000 times the cost rate. Under B it is 1,150 minus 1,250 times the cost rate. Set them equal and 250 times the rate equals 75, so the rate is 30 percent. Check it: at 30 percent, A gives 1,075 minus 300, which is 775, and B gives 1,150 minus 375, which is also 775.

Below a 30 percent cost of goods the discount wins on contribution. Above it the rewards program does. One sentence, and it converts a revenue answer into a decision a chief executive can use. We note in passing that the contribution gap at our assumed 50 percent rate is 50 million dollars, the same magnitude the firm publishes in the other direction, and we draw nothing from that coincidence.

Segmented prices exist in published data

Segmentation is the move most pricing guides recommend and none of them evidences. Here it is, published and checkable.

The Energy Information Administration publishes average US electricity prices by end-use sector at eia.gov, checked 30 September 2026. For July 2026 the national averages were 18.31 cents per kilowatt hour residential, 14.53 commercial, and 9.77 industrial, with all sectors together at 14.99. The agency labels the values preliminary.

Divide the residential price by the industrial: 1.874. The same commodity, the same month, the same country, sold at 87 percent more to a household than to a factory.

That is the shape of a segmented price, and it also shows the limit of the inference. The agency publishes prices and no cost breakdown, so the gap cannot be attributed to willingness to pay alone. Delivery costs differ by voltage, load shape and connection, and none of that is in the table. A candidate who cites a price gap as proof of segmentation value has skipped that check, and the honest version names both causes and says which data would separate them.

Useful consequence for a case: when you propose segmentation, propose the fence with it. Sector, voltage and contract length are the fences in that table. A segment with no fence is one price with extra steps.

Two firms put value in different places

Simon-Kucher is the pricing house in this set, and its recruiting page tells candidates something the other two do not. Michaela Scherhag, Team Lead Junior Recruiting in Frankfurt, writes at simon-kucher.com, checked 30 September 2026: "A solid understanding of economic principles and frameworks can also be beneficial, especially in our core focus areas such as pricing, marketing, digitalization, and sales."

Pricing is named first. So on this case type, the firm whose core practice it is tells you the standard apparatus is worth knowing, which is the opposite of what Kearney publishes about frameworks in general. Both positions are quoted with their sources on structured thinking.

The two published cases then place value differently. Oliver Wyman computes it first and uses it as a ceiling. Bain never mentions what the garment is worth to the shopper, and treats price purely as a lever on volume. Its own note against raising prices reads: "This is an unlikely assumption and requires more research to understand our elasticity of demand." So the option that needs an elasticity is dismissed for lacking one while the other two options get full arithmetic.

L.E.K. publishes a third prompt that puts the two halves together in one sentence: "Estimate the size of the market for this drug and describe how you would go about pricing this product." No model answer, which leaves the order to you. Compute value first and the size question becomes a volume check on a price you can already defend.

What the scale rewards in a pricing answer

The weights below are ours, read from the grading code in src/lib/grading. No consulting firm publishes a marking scheme, and the two cases above publish criteria per question with no weights attached.

Drilling price against volume in a pricing case

Pricing arithmetic is the most checkable in the case canon, because every answer has an inverse you can compute.

1. Invert the discount. A 20 percent price cut holds revenue only if volume rises by 1 divided by 0.8, so 25 percent. Compute that before you argue about whether the volume is plausible.
2. Find the pivot. Redo the FashionCo table at a 40 percent cost of goods. A gives 1,075 minus 400, which is 675, and B gives 1,150 minus 500, which is 650. A still wins, as it must above 30 percent.
3. Solve the coverage backwards. At what price does the Oliver Wyman product cover fixed cost at exactly half the target base? Half of 1.05 million is 525,000, and 10 billion divided by 525,000 is 19,048 rupees.

Run the arithmetic against a clock in the math drill, which returns Answer, Working, Units and Interpretation separately with the first correction named. Both that and the breakeven calculator are free and need no account. Neither will choose a base for your price, and neither knows what your buyer thinks the product is worth.

For prompts where the objective is reach rather than margin, see ESG cases and non-profit cases.