Van Westendorp calculator: price sensitivity meter analysis

The price sensitivity meter (PSM) was introduced by the Dutch economist Peter van Westendorp in 1976. Instead of asking people what they would pay — a question they answer badly — it asks four questions about where the price stops making sense, and reads the answers as cumulative distributions.

Each curve answers one question at every price. Where “too cheap” and “too expensive” cross, the smallest possible number of people reject the price for either reason — that is the optimal price point, and “optimal” here means least rejected, not most profitable. Where “cheap” and “expensive” cross, as many people call the price a bargain as call it expensive; this indifference price point often lands close to the price of the category leader, because that is the price people have been trained by.

Below PMC the price is cheap enough that doubts about quality outweigh the bargain; above PME the sense that it is expensive outweighs the value. The stretch between them is the range a price can sensibly live in — a range, not a recommendation.

Why the raw answers matter

Most “Van Westendorp calculators” on the web take four numbers and draw four straight lines. That is not a price sensitivity meter: with one respondent there is no distribution, and averaging the four questions first throws away exactly the disagreement the method exists to measure. This page takes every respondent’s four answers, so the curves have the shape your sample actually has, including the flat stretches where no crossing exists.

PSM is cheap to run and easy to over-read. These are the ways it goes wrong, and they matter more than the four numbers it produces.

• It measures price perception, not demand. Nobody in the survey was asked whether they would buy. A price inside the acceptable range can still sell nothing, because the range says the price is not absurd, not that the product is wanted. If you need quantities, you need a method that asks about purchase — Gabor–Granger for a demand curve, or the Newton–Miller–Smith extension, which adds two purchase-intent questions to the same four prices.
• OPP is not the revenue-maximising price. It minimises outright rejection. Charging the OPP is a good way to leave money with customers who would have paid more. Most teams that use PSM well treat OPP as a floor for discussion and IPP as a reference to the market, then price above both and check retention.
• Answers anchor on whatever you showed. The description, the competitor you named, the currency and even the order of the questions move the answers. Two surveys about the same product with different framing produce different curves, and the curves look equally confident.
• Stated prices are not paid prices. People systematically say they would pay less than they do for a product they can see, and more than they do for one they have to imagine. PSM tells you about the shape of the disagreement, not the level.
• A mixed sample still produces a confident-looking answer. Put a startup buyer and an enterprise buyer in one data set and the curves flatten out: the crossings land in the empty gap between the two groups, the range gets very wide, and the OPP becomes a stretch rather than a point — which this page says out loud instead of printing a number. A single OPP across two markets is an average nobody wants to pay. The segment table exists for exactly this.
• An inverted range means the screen is off, not that the market is strange. While every response satisfies too cheap ≤ cheap ≤ expensive ≤ too expensive, the “cheap” curve is always above “too cheap” and “too expensive” always below “expensive”, so PME can never land below PMC. If you see an inverted range, you are keeping inconsistent responses — which is a finding about the questionnaire, not about the price.
• The consistency screen biases the sample. Responses whose four prices are out of order have to be dropped, but the people who answer inconsistently are usually the ones least familiar with the category — which is exactly the group a new product needs to reach. Watch how many rows you lose, and treat losing more than about 15% as a questionnaire problem.
• Small samples move a lot. Each respondent is worth a full step in four curves. With 20 answers a single person can shift the OPP by a double-digit percentage; this page warns below 30 and flags segments below 15.
• It knows nothing about your costs or your competitors’ moves. The acceptable range can sit entirely below your cost to serve. PSM is one input to a pricing decision, next to unit economics, packaging and what happens when a competitor cuts price.

Used carefully, PSM is good at one thing: showing you how much your market disagrees, and where the disagreement splits. That is usually a packaging answer — plans, seats, usage tiers — rather than a single number.

• Price into the top half of the range. Above IPP, below PME, then watch trial-to-paid conversion and churn rather than the survey. The trial-to-paid conversion calculator shows what a price change has to do to revenue to be worth it.
• Turn segment disagreement into packaging. If the segment table shows three ranges that barely overlap, that is three plans, not one compromise price. Rerun the analysis per segment with 50+ responses each.
• Check the unit economics of the price you picked. A defensible price still has to pay back acquisition: see the CAC payback period calculator and the LTV:CAC calculator.
• Ask about features separately. Willingness to pay is shaped by what is in the plan. The Kano model survey analyzer tells you which features are must-haves and which are delighters worth putting on a higher tier.
• Size the survey before you run it. The survey sample size calculator gives the number of responses you need per segment for the margin of error you can live with.

1. Van Westendorp P. “NSS Price Sensitivity Meter (PSM) — A new approach to study consumer perception of prices.” ESOMAR Congress, Venice, 1976 — the original four questions and the four intersections.
2. Newton D., Miller J., Smith P. “A market acceptance extension to traditional price sensitivity measurement.” American Marketing Association Advanced Research Techniques Forum, 1993 — the purchase-intent extension that turns PSM into a revenue estimate; it needs two questions this tool’s four-price input does not contain, which is why this page does not pretend to compute it.
3. Lipovetsky S., Magnan S., Zanetti-Polzi A. “Pricing models in marketing research.” Intelligent Information Management 3(5), 2011 — review of PSM and Gabor–Granger, including the criticism that stated prices overstate sensitivity.
4. Harmon R., Unni R., Anderson T. “Price sensitivity measurement and new product pricing.” PICMET, 2007 — PSM applied to technology products, and the limits of the acceptable range.