The Van Westendorp price sensitivity meter
Four questions, four curves, four price points. Paste your survey responses below and the calculator plots them — then read what the method can and cannot tell you.
Short answer
What is the Van Westendorp price sensitivity meter?
A four-question survey that maps how a price is perceived. Each respondent names the price at which your product is too cheap, cheap, expensive, and too expensive. Plot the answers cumulatively and the crossings give you a floor, a ceiling, and two candidate prices in between.
The short version
- Published by Dutch economist Peter van Westendorp in 1976, and still the most widely used pricing survey in product research.
- Four crossings: PMC (floor), IPP, OPP (fewest rejections), PME (ceiling). PMC to PME is the range of acceptable prices.
- It measures perception, not purchase. There is no demand curve and no revenue estimate in the classic four questions.
- Responses whose four prices are not in ascending order are discarded — a sample where many are discarded means the wording was unclear.
Free · No signup · Nothing leaves your browser
Paste your four columns, get the four points
One row per respondent, four prices in the order of the four questions: too cheap, cheap, expensive, too expensive. Paste straight from a spreadsheet — tabs, commas, semicolons and spaces all work, and currency symbols are stripped.
Example data: 15 responses to a fictional SaaS pricing survey. One of them has its prices out of order on purpose, so you can see what the tool does with it.
14responses used
1discarded — prices not in ascending order
Point of marginal cheapness
Below this, more people suspect the quality than think it is a bargain. The floor.
Indifference price point
As many call it cheap as call it expensive. Often where the market leader already sits.
Optimal price point
The fewest people reject it outright — for being suspiciously cheap or flatly unaffordable.
Point of marginal expensiveness
Above this, more people drop out than see it as merely expensive. The ceiling.
Range of acceptable prices: $18 – $30. Outside it, the share of people rejecting the price — for being suspiciously cheap at one end, plainly unaffordable at the other — rises faster than the share still considering it.
# PRICE SENSITIVITY METER: [What you are pricing] Method: Van Westendorp (four questions, cumulative curves) Date: 2026-09-11 Responses used: 14 Responses discarded (prices not in ascending order): 1 Point of marginal cheapness (PMC): $18 Indifference price point (IPP): $23 Optimal price point (OPP): $25 Point of marginal expensiveness (PME): $30 Range of acceptable prices: $18 – $30 ## What this does NOT say These numbers describe how the price is PERCEIVED. Nobody in this sample bought anything, so none of it measures willingness to pay, demand volume, or revenue. Treat the range as the opening question for customer conversations, not as an answer. ## Next [Who did you ask, and are they the people who would actually buy?] [What happens at $25 when you put it in front of a real buyer?] Calculated with the free Van Westendorp tool at https://gonogo.team/van-westendorp
What this cannot know. Every number above comes from people naming prices, not paying them. The method measures how a price is perceived: it says nothing about how many would buy, what revenue you would make, or whether the people you asked are the people who would actually pay. A price that survives the survey and dies in front of a real buyer is the normal outcome, not a broken calculation.
The range is where the conversation starts. A 15-minute session puts “will they pay” — one of the three critical criteria in our Go/No-Go rubric — under evidence instead of under a survey.
The four questions
Ask all four, in this order, about a product the respondent has just had described to them in concrete terms. Keep the unit the same in every question — per month, per seat, one-off — or the four answers will not be comparable.
- 1
Too cheap
At what price would this be so low that you would question the quality?
- 2
Cheap
At what price would this be a bargain — a great buy for the money?
- 3
Expensive
At what price would this start to seem expensive — not out of the question, but you would have to think about it?
- 4
Too expensive
At what price would this be so expensive that you would not consider buying it?
The four answers of one respondent make one row in the calculator above, in exactly this order.
Where the four points come from
Each of the four answers becomes a cumulative curve. Two of them run downwards: the higher the price, the fewer people still call it too cheap or cheap. Two run upwards: the higher the price, the more people call it expensive or too expensive. Four curves, drawn over the same price axis, cross each other in four places — and each crossing has a name.
| Point | Crossing of | What it means |
|---|---|---|
| PMCPoint of marginal cheapness | “Too cheap” × “Expensive” | The floor. Below it, more people doubt the quality than see a bargain. |
| IPPIndifference price point | “Cheap” × “Expensive” | As many call it cheap as call it expensive. Often where the established brand already sits. |
| OPPOptimal price point | “Too cheap” × “Too expensive” | The fewest outright rejections. Equal numbers walk away for opposite reasons. |
| PMEPoint of marginal expensiveness | “Cheap” × “Too expensive” | The ceiling. Above it, more people drop out than merely grumble. |
PMC to PME is the range of acceptable prices. Price below it and a growing share of people conclude something must be wrong with the product; price above it and a growing share stop considering it at all. The calculator shades that range on the chart.
What Van Westendorp cannot tell you
It does not measure willingness to pay. Nobody in the survey bought anything. People name prices freely when no money is at stake, and they name them differently when a card is on the table. The four points describe perception, which correlates with purchase but is not the same thing.
There is no demand and no revenue in it. The method produces no volume estimate, so it cannot tell you which price earns most. The Newton-Miller-Smith extension bolts two purchase-intent questions onto the four in order to estimate trial and revenue — worth knowing that it exists, and worth knowing that the classic four questions are not it.
It is only as good as who you asked. The arithmetic cannot distinguish a qualified buyer from a curious passer-by, and it will happily produce a confident-looking range from the opinions of people who would never buy.
Averaging segments hides the answer. Two groups with genuinely different willingness to pay produce curves that cross in a place neither group would accept. Analyse them separately, or you get a price that satisfies nobody.
None of this makes the method useless. It makes it a starting point: a defensible range to test, and a much better opening than a number picked because a competitor charges it.
Four ways the survey goes wrong
❌ Asking people who would never buy
The method has no way to tell a prospective customer from a passer-by. Ask a general panel and you measure what the general public thinks your product should cost, which is a different and useless question.
✓ Instead: Screen first: category users, recent buyers, people with the problem. A hundred strangers are worth less than thirty qualified respondents.
❌ Reading the optimal price point as “the price”
OPP is the price with the fewest outright rejections. Fewest rejections is not most profit — a price that nobody objects to is frequently a price that leaves money on the table.
✓ Instead: Treat the range as the search space and the OPP as its centre of gravity. Profit maximisation needs volume and margin, which this survey does not measure.
❌ Running it on a product people cannot picture
Respondents price what they imagine. Without a concrete description — what it does, what it replaces, who it is for — the four answers are anchored to nothing and the curves wander.
✓ Instead: Show the same one-paragraph description, the same feature list and the same unit (per seat, per month, one-off) to everyone before the four questions.
❌ Quietly dropping the inconsistent responses
Some people answer with the prices out of order. Standard practice discards those responses — but if you discard them silently, nobody can tell a clean sample from one where half the respondents misread the questions.
✓ Instead: Report both numbers: how many you used and how many you dropped. The tool on this page shows the discarded count next to the accepted one for exactly this reason.
After the range, the conversation
A price range is an answer to “what would people accept?” The harder question sits one step behind it: will these people pay at all? In our Go/No-Go rubric that is one of three critical criteria — the ones no amount of strength elsewhere can compensate for.
A 15-minute voice session works through your idea across all seven criteria, including that one, and ends with a written GO / WAIT / NO-GO and the reasoning behind it. The range you just calculated is a good thing to bring into it.
15 min · free tier, no card
Frequently asked questions
What is the Van Westendorp price sensitivity meter?+
How many responses do I need?+
What do PMC, IPP, OPP and PME mean?+
Does Van Westendorp tell me how much revenue I will make?+
Why did the tool discard some of my responses?+
Is this calculator free, and does my data leave the browser?+
Related guides