Van Westendorp Price Sensitivity Meter

The four Van Westendorp questions

The Price Sensitivity Meter (PSM) was introduced by the Dutch economist Peter van Westendorp in 1976. After showing a product or concept, the survey asks four open-ended price questions:

1. Too cheap: At what price would you consider the product to be priced so low that you would feel the quality could not be very good?
2. Cheap (a bargain): At what price would you consider the product to be a bargain, a great buy for the money?
3. Expensive: At what price would you say the product is starting to get expensive, so that it is not out of the question, but you would have to give some thought to buying it?
4. Too expensive: At what price would you consider the product to be so expensive that you would not consider buying it?

Each respondent therefore gives four prices that should increase from too cheap to too expensive. The method is popular because it needs no prior knowledge of the right price range, works for new products and categories, and is quick to field.

How the curves are built

For every possible price p, the tool calculates four cumulative percentages:

• Too cheap(p): share of respondents whose “too cheap” price is at or above p. This curve falls as price rises: almost everyone thinks a very low price is suspiciously cheap.
• Cheap(p): share whose “bargain” price is at or above p (falling).
• Expensive(p): share whose “getting expensive” price is at or below p (rising).
• Too expensive(p): share whose “too expensive” price is at or below p (rising).

The classic analysis also uses the complements Not cheap = 1 − Cheap and Not expensive = 1 − Expensive. The four key price points are where pairs of curves cross; the tool finds each crossing on a fine grid of 800 prices with linear interpolation.

The four price points

• Point of marginal cheapness (PMC): Too cheap crosses Not cheap. Below this price, more people doubt the quality than see a bargain. It is the lower bound of the acceptable range.
• Point of marginal expensiveness (PME): Too expensive crosses Not expensive. Above it, too many people reject the price. It is the upper bound.
• Optimal price point (OPP): Too cheap crosses Too expensive, the price at which the fewest people reject the product for either reason.
• Indifference price point (IPP): Cheap crosses Expensive. As many people think the price is a bargain as think it is getting expensive; it is often interpreted as the normal or expected market price.

Some textbooks and software use a simplified convention in which the PMC is where Too cheap crosses Expensive, and the PME where Too expensive crosses Cheap. Both conventions are available in the options; the classic one is the default. State which you used when you report the results, because the bounds can differ by a few percent.

The Newton-Miller-Smith extension

The basic PSM shows acceptability, not demand: it does not tell you how many people would actually buy at each price. Newton, Miller and Smith (1993) added two purchase-likelihood questions, asked at the respondent’s own “cheap” and “expensive” prices, usually on a five-point scale. This tool calibrates those answers to probabilities (5 = 70%, 4 = 50%, 3 = 30%, 2 = 10%, 1 = 0%, the calibration in the original paper), sets purchase probability to zero at each respondent’s “too cheap” and “too expensive” prices, interpolates between the four points, and averages across respondents. The result is an expected trial curve and a revenue curve (price × trial), whose maximum is a revenue-optimal price. Add the two columns in the variable list to get this analysis.

Worked example with the built-in data

The example contains 260 simulated respondents evaluating a product with a typical price of around 25. About 6.5% gave inconsistent answers (prices not in increasing order) and were excluded, which is a normal rate for this kind of question. The acceptable price range runs from about 16 to about 34, the OPP and IPP both lie near 24, and with the purchase-likelihood questions, expected revenue peaks at a price of about 25, where roughly 23% of respondents are expected to buy. A sensible launch price would be in the 24–26 band, close to the optimal and indifference points and at the revenue peak.

Cleaning the data

Before running the analysis, check that every answer is a number in the same currency, remove obvious typing errors (an extra zero, a price in cents) and decide what to do with inconsistent respondents. The standard practice, and the default here, is to exclude anyone whose four prices are not in non-decreasing order. Extreme answers barely affect the crossings, because the curves are based on ranks, but the chart range is trimmed to the bulk of the answers so that a single absurd price does not squash the plot.

How to use the results in pricing decisions

Treat the PMC–PME range as the window in which the product is acceptable to the market, and the OPP–IPP region as the zone where a launch price is least likely to provoke resistance. A narrow range signals a price-sensitive market with clear expectations; a wide range suggests room to differentiate on price. If the IPP is well above the OPP, customers expect to pay more than the price that minimises rejection, which supports a premium position. Then validate a short list of prices inside the range with Gabor-Granger (direct buy/no-buy questions) or choice-based conjoint (prices in competitive context).

Limitations

• Respondents state prices without seeing competitors; the PSM measures perception, not choice behaviour.
• Answers anchor on the price people already pay; for established categories the results often reproduce the current market price.
• The method does not estimate elasticity or volume without the NMS extension, and even then purchase intent is stated, not observed.
• Segments can differ sharply; run the analysis separately for key segments (for example current customers and prospects) when sample size allows, at least 100 per group.

Reporting

Report the sample size and the number of excluded respondents, the convention used for the intersections, the four price points, and the chart. Example: “A Van Westendorp analysis (N = 243 after excluding 17 inconsistent respondents) indicated an acceptable price range of 16.05 to 34.02, an optimal price point of 24.02 and an indifference price point of 23.98.”

Key references: Van Westendorp (1976), ESOMAR Congress; Newton, Miller and Smith (1993), ESOMAR; Lipovetsky, Magnan and Zanetti-Polzi (2011), Journal of Data Science.