Market research tool
Van Westendorp price sensitivity meter
How results are checked
Named statistical review pending. Selected automated reference cases are published on the validation page.Range of acceptable prices
20.91–39.65per month
Optimal price point 29.50 · indifference price point 29.85
NMS revenue peaks at 31.00 · 23 out-of-order respondents excluded
The Van Westendorp Price Sensitivity Meter asks four open price questions — at what price is the product too cheap to trust, a bargain, getting expensive, and too expensive to consider. Plotting the cumulative share of answers at each price gives four curves; where they cross marks the range of acceptable prices (from the point of marginal cheapness to the point of marginal expensiveness) and the optimal and indifference price points.
Run a Van Westendorp analysis on your data
Open a respondent-level file with one column per price question, pick the four columns, and the curves and price points appear at once. Optionally add a weight and the two Newton-Miller-Smith purchase-likelihood questions. Everything runs in this browser tab — free, no account, and the file is never uploaded. "Use sample data" loads a synthetic study of 400 respondents pricing a monthly meal-kit subscription.
CSV, Excel, SPSS, Stata, SAS, R, Parquet, SQLite or JSON. The file is read in this browser tab and never uploaded. For Triple-S, select the .sss or .xml metadata file and its data file together.
How the curves and intersections are computed
The calculator evaluates each cumulative curve at every distinct price a respondent typed: at price p, "too cheap" is the share whose too-cheap answer is at or above p, "cheap" the share whose cheap answer is at or above p, "expensive" the share whose expensive answer is at or below p, and "too expensive" the share whose too-expensive answer is at or below p. "Not cheap" and "not expensive" are their complements.
Neighbouring prices are joined by straight lines and each intersection is solved on those lines (linear interpolation), so a price point can fall between two answers. If two curves are exactly equal over several consecutive prices, the midpoint of that run is reported. Respondents who give the four answers out of order (too cheap ≤ cheap ≤ expensive ≤ too expensive, equal answers allowed) are counted and, by default, excluded. For crosstabs of the price answers by segment, use the survey crosstab tool.
The method
Formula
PMC: TooCheap(p) = NotCheap(p) · OPP: TooCheap(p) = TooExpensive(p) · IPP: Cheap(p) = Expensive(p) · PME: TooExpensive(p) = NotExpensive(p)
where
- TooCheap(p)
- Σ wᵢ · 1[too-cheap answerᵢ ≥ p] / Σ wᵢ (weights wᵢ = 1 when unweighted)
- Cheap(p)
- share whose cheap answer is ≥ p; NotCheap(p) = 1 − Cheap(p)
- Expensive(p)
- share whose expensive answer is ≤ p; NotExpensive(p) = 1 − Expensive(p)
- TooExpensive(p)
- share whose too-expensive answer is ≤ p
- NMS Trial(p)
- Σ wᵢ · πᵢ(p) / Σ wᵢ, where πᵢ is 0 at the respondent's too-cheap and too-expensive prices, the calibrated likelihood at their cheap and expensive prices, and linear in between; revenue = p × Trial(p)
Worked example
In the synthetic sample, 23 of 400 respondents answered the four questions out of order and are excluded, leaving 377. The point of marginal cheapness is 20.91 and the point of marginal expensiveness 39.65, so the range of acceptable prices runs from 20.91 to 39.65 a month.
Inside that range the optimal price point (too cheap = too expensive) is 29.50, where 9.4% find the price too cheap and the same share too expensive, and the indifference price point (cheap = expensive) is 29.85. Keeping the out-of-order respondents moves the optimal price point to 29.41.
With the Newton-Miller-Smith questions and the default calibration (0, .1, .3, .5, .7 for answers 1–5), expected revenue per respondent peaks at 31.00 with 23.9% trial; trial itself is highest at 25.00 (26.2%).
Using it well
When to use it
Use it when
- Use: Early pricing research for a new product or service, before you know which prices to test.
- Use: Finding the band of prices buyers find credible — neither suspiciously cheap nor out of reach.
- Use: You asked the four Van Westendorp questions as open numeric answers for every respondent.
Not the right tool when
- Avoid: You need demand at specific prices or a price against competitors — use Gabor-Granger or a choice-based conjoint.
- Avoid: The category has a strong reference price (respondents anchor on it and the curves collapse).
- Avoid: Respondents cannot judge the price without seeing the product; PSM measures perception, not purchase.
How to interpret it
Rule of thumb
Treat the acceptable range as the realistic pricing window, not a demand forecast. A narrow range means buyers share a clear idea of value; a wide one leaves room for positioning. The optimal price point is where the fewest people reject the price for being too cheap or too expensive — it is not the revenue-maximising price. For revenue, add the Newton-Miller-Smith questions or test specific prices with Gabor-Granger.
Questions
Frequently asked questions
What are the four Van Westendorp questions?
At what price would you consider the product to be so inexpensive that you would doubt its quality (too cheap)? At what price would it be a bargain — a great buy for the money (cheap)? At what price would it start to seem expensive, though you would still consider it (expensive)? At what price would it be so expensive that you would not consider buying it (too expensive)?
Which curves define PMC and PME?
This calculator uses the classic definitions: the point of marginal cheapness is where 'too cheap' crosses 'not cheap', and the point of marginal expensiveness is where 'too expensive' crosses 'not expensive'. Some practitioners use 'too cheap' × 'expensive' and 'too expensive' × 'cheap' instead, which gives a slightly different range.
How are the intersections interpolated?
Each curve is evaluated at every distinct price respondents gave and joined with straight lines; an intersection is solved on those lines, so it can fall between two answered prices. If two curves are exactly equal over consecutive prices, the midpoint of that run is reported.
What happens to respondents who answer out of order?
A respondent is consistent when too cheap ≤ cheap ≤ expensive ≤ too expensive. Out-of-order respondents are counted and excluded by default; untick the option to keep them. Respondents missing any of the four answers are always left out.
What does the Newton-Miller-Smith extension add?
Two follow-up questions ask how likely the respondent is to buy at their own cheap and expensive prices. Each answer is calibrated to a probability (by default 70%, 50%, 30%, 10% and 0% from 'definitely' to 'definitely not'), interpolated across prices, and averaged into a trial curve. Price × trial gives a revenue curve whose peak is the revenue-maximising price. Where a respondent's answers coincide (say too cheap = cheap) their probability jumps at that price; the calculator reads trial just below and just above each such price rather than at the single price itself, so a maximum is always a level the curve holds over a stretch of prices, not a one-price spike.
Sources and next steps
References & further reading
- [1]Van Westendorp, P. (1976). NSS Price Sensitivity Meter (PSM) — A new approach to study consumer perception of prices. Proceedings of the ESOMAR Congress, Venice.
- [2]Newton, D., Miller, J., & Smith, P. (1993). A market acceptance extension to traditional price sensitivity measurement. Proceedings of the American Marketing Association Advanced Research Techniques Forum.
Try it on your own data — free, no signup
Upload a CSV, Excel, SPSS, Stata, SAS, R, Parquet or JSON file. Everything runs in your browser; your file never leaves your device.
Open the workspace →