Van Westendorp price-perception analyzer
What prices feel reasonable to your audience?
Turn four price judgments per respondent into an interpretable range. Review inconsistent answers, compare calculation conventions, and see how stable the intersections are.
This tool analyzes stated price perceptions. Choosing a selling price also requires evidence about demand, competition, and costs.
- Too cheapThe price raises doubts about quality or credibility.
- A bargainThe offer feels inexpensive and good value.
- ExpensiveThe price is high, but the offer remains worth considering.
- Too expensiveThe price rules the offer out.
Define one offer before asking about price.
Specify the configuration, quantity, billing period, currency, and treatment of fees or taxes. Respondents need to judge the same offer and use the same price unit. Record the actual wording and question order with your analysis.
A four-question survey starting point
After describing the offer, ask each person for four numeric amounts in the stated currency. Adapt and pretest this wording for your product:
- How low would the price need to be before you doubted the offer’s quality?
- What price would make the offer feel like a bargain?
- At what price would the offer start to feel expensive while still being worth considering?
- What price would make you rule out buying the offer?
Save the answers as too_cheap, cheap, expensive, and too_expensive. An anonymous respondent_id and a segment column are optional. The downloadable demonstration provides a file template, including deliberately problematic rows.
When a different pricing study fits better
Van Westendorp summarizes perceptions of a defined offer. A Gabor–Granger or monadic study asks about response at specified prices. Conjoint research examines choices involving features, prices, and competitors. None turns hypothetical answers into guaranteed sales. See how pricing research methods fit the decision.
Read the assumptions with the result.
All four curves share the same complete analytical base after the selected ordering and identity rules. The default requires too cheap < cheap < expensive < too expensive. Equal and reversed answers remain visible in diagnostics and sensitivity comparisons.
What do the four intersections mean?
The traditional Optimal Price Point (OPP) crosses “too cheap” with “too expensive.” The Indifference Price Point (IDP) crosses “cheap” with “expensive.” Under the original range definition, the Point of Marginal Cheapness (PMC) crosses “too cheap” with “not cheap”; the Point of Marginal Expensiveness (PME) crosses “not expensive” with “too expensive.”
The narrower convention uses “too cheap” with “expensive” for PMC, and “cheap” with “too expensive” for PME. The acceptable range runs from PMC to PME; unavailable or reversed endpoints withhold it. These names describe conventional intersections and do not establish a financial optimum.
How are ties, steps, and gaps handled?
At each distinct observed price p, the empirical cumulative fraction counts thresholds ≤ p. “Expensive” and “too expensive” use that fraction; “cheap” and “too cheap” use its complement. Thus answers exactly at p count on the cumulative side. No rounding is applied before calculation.
Reported crossings use straight lines between adjacent observed-price knots. A coincident interval uses its midpoint, with both endpoints retained. The empirical-step chart and observed brackets make gaps visible. This explicit endpoint and interpolation policy may differ from other software defaults; it is included in every methods receipt. A bracket is not an uncertainty interval.
What does resampling establish?
The optional analysis samples whole usable respondents with replacement 500 times, recomputing the crossings with a fixed seed. It reports the 2.5th and 97.5th percentiles, plus failed and coincident crossings. Results are conditional on the selected cleaned base; they do not measure uncertainty from excluding records.
Resampling assumes independent respondents. The intervals describe stability within this sample and do not establish population coverage for an opt-in or convenience sample. Survey weights, stratification, clusters, paired segment differences, and Newton–Miller–Smith purchase-intent modeling are outside this release.
Where do imported records go?
Import, validation, calculations, and downloads run in your browser. The analyzer does not send respondent records to Russell, place them in contact links, or save them between visits. Clear data removes the current analysis from the tool. Your original file and any downloads stay on your device.
XLSX import reads the first worksheet and does not recalculate formulas. Use values-only cells. Both text and spreadsheet imports require one header row, at most 5,000 respondent rows, and prices from zero to one billion. These are software limits, not research recommendations.
Sources
- pricesensitivitymeter: package reference. Documents threshold validation, the original and narrower range definitions, and intersection choices. This browser implementation uses the explicit numerical policy above.
- Sawtooth Software: Van Westendorp pricing sensitivity meter. Explains the survey questions, perceptual interpretation, and limits of using conventional intersections to set prices.
- AAPOR: Transparency Initiative. Supports disclosure of recruitment, sample construction, questions, processing, and analysis assumptions. It does not certify this calculator or its resampling intervals.