Pricing cases: how to set, raise and defend a price

What is a pricing case interview, and what kinds are there?

A pricing case hands you one decision: what should this product cost? The client might be launching something new, reacting to a rival, or sitting on a price nobody has touched for years. Whichever it is, the interviewer wants to see whether you can hold three things in your head together, which are the customer's view of value, the competitive picture and the unit economics, and then do the arithmetic that says whether a move pays.

It is a close cousin of the profitability case. Profitability asks where the profit went. Pricing zooms into one branch of that tree, price, and adds something the tree cannot: behaviour. Change the price and customers, competitors and your own sales channel all react, and a good answer has a view on each reaction.

The prompts arrive in five recurring shapes. Recognise which one you have been given in the first thirty seconds, because it tells you which lens is likely to decide the case.

Pricing also turns up as a single question inside other cases. A market-entry case usually ends with "what would we charge, and does the profit work?" A private-equity deal screen often asks whether the target has pricing power it has not used.

One note on format. Some pricing cases are candidate-led, where you build the structure and pick the analysis. Others are interviewer-led, where the interviewer poses a fixed sequence of questions and hands you the data as you go. The guide to interviewer-led and candidate-led cases covers how to adapt. Pricing suits the interviewer-led format well: structure, one maths step, then a recommendation.

What are the three lenses, and which one decides the answer?

Every price you will ever be asked about can be looked at from three sides. Cost-based pricing starts from what the product costs to make and adds a margin. Competitor-based pricing starts from what the nearest alternatives sell for. Value-based pricing starts from what the product is worth to the customer, and prices somewhere below that.

BCG's article on strategic pricing (2023) puts the same three inputs, cost, competition and customer value, at the centre and argues that which one should lead depends on the kind of market you are in: it names a cost game, a power game, a value game and a uniform game. We are not affiliated with BCG and we use their framing lightly here, but the idea travels well into a case room. In the room, ask yourself which lens decides the answer for this client and where the other two set the boundaries; asking which lens is right in general gets you nowhere.

Why is cost-plus alone a weak answer? Because cost tells you what you must charge to break even, and the customer has never heard of your cost. Take a product that costs £20 to make. Cost-plus at 50% gives £30. If customers would happily pay £60, you have left £30 on every unit. If they would only pay £25, you have priced yourself out and will sell nothing. The number is wrong in both directions for the same reason: it contains no information about the buyer.

Harvard Business Review's 2016 guide to value-based pricing by Utpal Dholakia makes a related point: value-based pricing is widely discussed and widely misunderstood, and companies that fall back on cost-based methods leave money on the table. We would add a caution from our side of the table. Candidates who have read that argument tend to say "we should price on value" and stop. Interviewers give no credit for the label. They give credit for the estimate of value, the evidence behind it and the arithmetic that follows.

Here is a quick test of which lens leads.

How do you structure a pricing case?

Do not open with the three lenses as if they were a recited list. Build the structure around the decision, and let the lenses appear where they belong. A pricing tree needs four branches: the customer, the competitors, the economics, and the risks of acting. The first two set the range in which the price can sit, the third tests whether a specific move pays, and the fourth is where recommendations get stress-tested. The habit behind good trees is covered in MECE and issue trees.

The order of your questions in the room counts for more than the tree. Here is the sequence we would follow.

How does the price-volume maths work?

The maths is one line of algebra, and once you can derive it you never need to memorise it. Start with contribution margin: the price minus the variable cost of one unit, or that amount as a share of the price. A £100 product with £60 of variable cost has a contribution of £40 a unit, or 40%. Fixed costs are ignored on purpose. They do not change when the price changes, so they drop out of the question of whether a price move helps.

Now raise the price and ask how many units you can lose before you are worse off. You are worse off only when total contribution falls, so set the new contribution equal to the old one and solve.

So at a 40% contribution margin a 10% price rise can lose up to a fifth of your volume before profit falls. Most people expect a smaller number, and it is why price rises so often pay. Notice what happens to revenue at that breakeven: 800 units at £110 is £88,000, against £100,000 before. Revenue is down 12% and profit is unchanged. A client watching revenue alone will call that a failure, so say the word contribution first.

The same logic runs the other way for a cut. Lowering the price by a fraction p needs a volume gain of p ÷ (M − p) just to stand still. At 40% margin, a 10% cut needs a third more units (0.10 ÷ 0.30 = 33.3%). Check it: 1,333 units at £90, margin £30, is £40,000 again. Cuts are much more expensive than rises are cheap, and the thinner the margin the sharper the asymmetry.

Read the table down the columns. High-margin businesses can tolerate less volume loss on a rise, because each lost unit takes more contribution with it, and they need less volume gain to pay for a cut. Thin-margin businesses have the opposite profile: they can lose a lot of volume on a rise and stay ahead, but a cut has to double the volume to break even at a 20% margin. A retailer on a 20% margin who discounts by 10% is asking its tills to ring twice as many sales.

The estimator below lets you run your own version. We set the defaults to the example: 10,000 units a year, a £100 product with £60 variable cost, so £40 contribution and £400,000 in total. After a 10% rise the new unit margin is £50, and we start the volume kept at 80%, which is the breakeven. Drag the volume kept above 80% and profit rises. Drag it below and it falls. Change the new margin to test a smaller or larger price move.

Two cautions before you use this in a room. First, the formula assumes variable cost per unit stays the same and that fixed cost does not change. If a big volume loss would let the client shut a line or drop a production line, say so, because then the true breakeven is a little more forgiving. Second, the number is a tolerance. It tells you how much volume you can lose. It does not tell you how much you will lose, and that second number is elasticity.

What is price elasticity, in plain words?

Price elasticity is the percentage change in volume for a one percent change in price. If a 10% price rise costs you 20% of your units, elasticity is minus two: for every percent on the price, two percent off the volume. BCG describes it in the same terms, as how demand for a product varies with its price, and treats it as the point where cost and customer value meet in setting an optimal price.

Put the breakeven and elasticity side by side and the case answers itself. The breakeven says you can lose 20%. Elasticity is your estimate of what you will lose. If you expect to lose 8%, the rise is comfortably profitable. If the evidence says 25%, it is not. The recommendation is the gap between the two numbers, and you should say the gap out loud.

Where does the number come from? In a case, the best answer is the history of the client's own prices. Ask whether the price has moved before and what happened to volume afterwards, whether different regions or channels pay different prices, and whether promotions have been run. Any of these gives you a real estimate and beats a theory. If there is no history, say what you would test, such as a price trial in one region or one channel, and what result would change your mind.

Some things make demand less sensitive, and knowing them tells you where to push and where to hold.

• Few close substitutes, so the customer has nowhere obvious to go
• The product is a small share of what the customer spends in the category
• Switching costs: retraining, contracts, data, habit
• The price is read as a mark of quality, which is the case in luxury and in some professional services
• The buyer is not the payer, as with expense-account purchases and insured medical care

The opposite conditions make demand sensitive: commodity products, price comparison sites, and buyers who are spending their own money on something they could postpone. Elasticity also varies by customer segment, and by time. A rise that loses 5% of volume in the first quarter can lose 12% by the end of the year as customers find alternatives. A sharp candidate says both things: "I would expect the short-term loss to be smaller than the long-term loss, so I would check the breakeven against the long-term figure."

One habit to avoid: quoting an elasticity number for a category as if you had looked it up. You have not, and neither has the interviewer. Say what you would expect qualitatively, ask for evidence, and put a range on it. That is what the rubric means when it says the chain of reasoning should be auditable. The case interview maths guide covers the wider arithmetic toolkit.

How do you price a new product?

New products have no price history, no elasticity data and often no direct competitor, which is why cost-plus is so tempting and so poor here. The strong answer is value-based, and it has a specific shape: find the customer's best alternative, estimate what your product does better in money terms, and price between the two.

Practitioners call the pieces the reference value and the differentiation value. The reference value is what the customer pays today for the best alternative. The differentiation value is the money-equivalent of whatever your product does better, net of the trouble of switching. Add them together and you have the economic value to the customer. That is the ceiling. Nobody rational pays more than it, and you would not want them to pay all of it, because a customer who keeps none of the gain has no reason to change. Bain's pricing practice describes using value-based approaches for new innovations, and the sentence in the client meeting is the same as the one in the case room: price below the value, and leave the customer a reason to switch.

Here is an example we would be happy to see a candidate reason through. A software tool saves a warehouse team 300 hours a year, and an hour of that team's time is worth £40. The tool they use today costs £8,000 a year. The differentiation value is 300 × £40 = £12,000. If the vendor prices to capture 40% of that gain, the price is £8,000 plus 40% of £12,000, which is £8,000 + £4,800 = £12,800. The customer's total value is £20,000, so they keep £7,200 of surplus.

Two choices come up in almost every new-product case. The first is skimming versus penetration. Skimming means launching high and coming down as the early adopters are served. Penetration means launching low to win volume, a market position or scale. We lean towards skimming when the product is clearly differentiated, capacity is limited, or a lower price is hard to reverse, and towards penetration when there are network effects or when volume brings a real cost advantage. The second choice is whether one price is right at all. If customers differ a lot in what they would pay, two or three versions at different prices will make more than one price, provided the cheaper version does not cannibalise the dearer one.

What do you do when a competitor changes its price?

This is the prompt where candidates rush. A rival cuts by 10%, the client panics, and the candidate reaches for "match it" or "hold and differentiate" inside a minute. Neither is a good first move, because both skip the question of who is at risk.

Start with what the rival did and to whom. Is the cut on the product that competes with ours, or on a different line? Is it permanent, or a promotion? Is it aimed at our most valuable customers, or at a segment we do not care about? Then look at our own customers: which of them compare on price, which are locked in, and which value something the rival cannot supply? Only a subset of your base is truly at risk, and the response should be aimed at that subset.

Then do the maths on matching. At a 40% contribution margin, a 10% cut needs 33.3% more volume to break even, as the table showed. A rival's cut does not bring you extra volume, so matching across the whole range is almost always a loss. It costs you the full cut on all the units you would have kept anyway. That is the strongest single argument against matching, and it is one arithmetic line.

What is different about luxury and premium pricing?

In most goods the price is a cost to the buyer. In luxury it is also part of what they are buying. A handbag at a price everyone can afford is a different product from the same bag at a price few can, and the customer knows it. That changes the logic of every lens.

Start with value-based pricing. The customer's value includes the meaning of the brand: craft, heritage, what carrying it says about them. For a luxury bag the reference alternative is another luxury house's product, or no purchase at all; a cheaper bag with the same function barely enters the buyer's comparison. And the willingness to pay for the top of the range is often set by what the best-known pieces cost, so the iconic product carries the price image for the whole house.

Competitor-based pricing changes too. A premium brand does not aim for parity with the field. It aims to sit where its customers expect it to, relative to a small set of peers, and it holds that position over years. A rival's discount is often best ignored, because copying it would tell customers the brand is coming down to that level.

Then the maths. Luxury margins tend to be high, so from the table above the tolerated volume loss on a rise is small. A high-margin brand cannot lose much volume on a rise. But luxury demand is also often less sensitive to price than mass goods, and in some categories a higher price makes the item more desirable to some buyers. Both are true, and the candidate who says so, and then asks for evidence on which one wins for this product, is doing the job. Assuming either without evidence is the error.

Discounting deserves its own paragraph. A promotion in a mass market moves volume and ends. In luxury it teaches every customer that the price is negotiable, damages the resale value of what earlier customers bought, and pushes the brand towards a place from which it is hard to come back. We would treat markdowns as a last resort and prefer levers that protect the price: limited editions, made-to-order, product changes, and keeping stock scarce.

Three more things to raise in a premium case where the prompt allows. Look at the price ladder: entry pieces, mid-range and iconic products each play a different role, and a rise on one rung affects the others. Look at consistency across markets, because tourists and online shoppers compare prices between countries. And look at channel, because a brand that sells through its own stores, wholesale partners and online has three different price conversations to manage at once.

What does a strong pricing recommendation sound like?

A recommendation has four parts: a number, the reasoning behind it, the volume it can survive, and the risks with a plan for each. It should take about 45 seconds. Here is a candidate on a price-rise case, speaking the way people do under pressure.

Check the arithmetic inside that answer, because the interviewer will. At 8% and a 40% margin the breakeven loss is 0.08 ÷ (0.40 + 0.08) = 0.08 ÷ 0.48 = 16.7%. Test it with 1,000 units at £100 and a £40 margin: £40,000. At £108 the margin is £48, and 833 units × £48 is £40,000. It works. (The price-history claim is part of the imagined case, and in a real one you would have been given the data or you would ask for it.)

The answer never says "it depends". It commits to a number, shows the working, and puts a range and a review date around it, which is what our commercial-judgement skill rewards.

The three risks in the script are the ones we see most in real pricing decisions, and each deserves a sentence of its own in your answer. Channel conflict comes first: distributors and retailers hold stock at the old price, resent being surprised, and can respond by cutting your shelf space. Customer reaction is next, and it is not uniform. Loyal customers may accept a rise that price-driven ones will not, so segment before you predict. Competitor response completes the set. Ask whether they can follow, whether they would want to, and what you would do if they did the opposite. A recommendation that names all three and gives each a mitigation reads as senior. One that names none reads as a calculation.

Framing, commercial judgement, rigor and synthesis all show up in one pricing answer, as how case interviews are scored sets out, which is why interviewers with thirty minutes like the case type.

What mistakes do candidates make in pricing cases?

These are the ones we see most in the transcripts we mark. The full list of case-interview errors is in the mistakes guide. This is the pricing-specific part.

• Opening with cost-plus and stopping there. It is the mark of someone who has not thought about the customer, and interviewers move on within a minute
• Naming the three lenses without saying which one decides. The list is a start, and the choice is the answer
• Dividing the price rise by the old margin, so a 10% rise at 40% gives 25% instead of 20%. The units you keep earn the new margin
• Reporting the breakeven volume loss as the number of units you expect to lose. It is the most you can afford to lose. Your forecast comes from elasticity and evidence
• Using gross margin where contribution margin belongs, which gives the wrong breakeven whenever fixed costs are inside the gross margin
• Judging a rise by revenue. Profit can hold while revenue falls 12%, and a client who is told about the revenue drop first may not hear the rest
• Quoting an elasticity for the category as if it were a known figure
• Ending with a number and no risks, or with risks and no number
• Matching a competitor's cut across the range without asking who is at risk

Most of these come from one habit, which is grabbing a formula before deciding which lens the case needs. The candidates who do best say what they are trying to establish before they compute it. If you practise alone, record yourself and check whether your first minute contains a lens, an objective and a hypothesis, or only a definition. Our guide to practising case interviews alone covers how to do that without a partner.