Van Westendorp price sensitivity meter: questions, analysis and limits

The Van Westendorp price sensitivity meter is a four-question survey that asks people at what prices a product would seem too cheap, a bargain, getting expensive, and too expensive. You plot the answers as four cumulative curves and read an acceptable price range from where they cross. It tells you how people perceive price for one described product. It does not tell you how many will buy.

The method was introduced in 1976 by Dutch economist Peter van Westendorp, according to Wikipedia's summary of the technique. It is still common because it is quick to field and easy to explain.

The four Van Westendorp questions

The standard wording, as listed by Wikipedia and Sawtooth Software, is close to this:

Too cheap: At what price would you consider the product to be priced so low that you would feel the quality couldn't be very good?
Bargain: At what price would you consider the product to be a bargain, a great buy for the money?
Getting expensive: At what price would you consider the product starting to get expensive, so that it is not out of the question, but you would have to give some thought to buying it?
Too expensive: At what price would you consider the product to be so expensive that you would not consider buying it?

Three details make the answers usable:

Describe one concrete offer. Say what is included, who uses it and the price unit, for example "per site, per month." If people imagine different products, the curves blend them.
Mention the alternatives. Sawtooth notes the method lacks competitive context and recommends asking respondents to consider substitute products. A sentence such as "Today most teams use a spreadsheet or a general work-order tool" helps.
Check the order. Each person's answers should rise: too cheap < bargain < getting expensive < too expensive. Drop or follow up on answers that do not.

A worked example with ten answers

Suppose you are pricing a work-order app for facilities teams that manage several buildings, charged per site per month. You ask ten facilities managers the four questions. The answers below are hypothetical, and ten is far too few for a real study; it keeps the arithmetic visible.

Respondent	Too cheap	Bargain	Getting expensive	Too expensive
A	$20	$40	$60	$90
B	$30	$50	$100	$150
C	$25	$45	$70	$100
D	$60	$90	$140	$200
E	$35	$60	$90	$130
F	$40	$70	$110	$160
G	$15	$30	$50	$75
H	$70	$100	$160	$240
I	$30	$55	$80	$120
J	$50	$80	$120	$180

Build the four curves

For each candidate price, count the share of people for whom that price falls in each category:

Too cheap at $P: their "too cheap" answer is $P or higher. This curve falls as price rises.
Bargain at $P: their "bargain" answer is $P or higher. This also falls.
Getting expensive at $P: their "getting expensive" answer is $P or lower. This rises.
Too expensive at $P: their "too expensive" answer is $P or lower. This rises.

Flipping the first two curves so they fall is the standard practice described on Wikipedia; without it you would not get the four crossings. For the ten answers above:

Price	Too cheap	Bargain	Getting expensive	Too expensive
$40	40%	90%	0%	0%
$50	30%	70%	10%	0%
$60	20%	50%	20%	0%
$70	10%	40%	30%	0%
$80	0%	30%	40%	10%
$90	0%	20%	50%	20%
$100	0%	10%	60%	30%
$120	0%	0%	80%	40%

Check one row by hand. At $60, respondents D ($60) and H ($70) gave "too cheap" answers at or above $60, so 2 of 10, or 20%, would call $60 too cheap. Only A ($60) and G ($50) gave "getting expensive" answers at or below $60, so that column also reads 20%.

Read the intersections

The four crossings have conventional names. Their meaning is debated, so treat them as landmarks, not answers.

Point of marginal cheapness (too cheap crosses getting expensive): both are 20% at $60.
Optimal price point (too cheap crosses too expensive): too cheap is 10% at $70 and 0% at $80, while too expensive goes from 0% to 10%. They cross between $70 and $80.
Indifference price point (bargain crosses getting expensive): 40% vs 30% at $70, then 30% vs 40% at $80, so between $70 and $80.
Point of marginal expensiveness (bargain crosses too expensive): both are 20% at $90.

The conventional reading is an acceptable range of roughly $60 to $90 per site per month, with $70 to $80 in the middle. In a real study you would interpolate on a finer price grid and use far more respondents; with ten people, one answer moves a curve by 10 points.

Sawtooth is blunt about the chart: the intersection names sound authoritative, but the analysis "doesn't follow from economic theory." Read Sawtooth's criticisms.

Add purchase intent with the Newton-Miller-Smith extension

The plain method says nothing about volume. The 1993 Newton-Miller-Smith extension adds two questions: how likely the person is to buy at their own "bargain" price and at their own "getting expensive" price, on a five-point scale. You then estimate purchase likelihood at each price and look for the price with the most expected revenue.

Because people overstate intent, analysts discount the ratings. Sawtooth gives one possible set: count 75% of "definitely would buy," 25% of "probably would buy" and 10% of "might or might not buy" answers as real purchases. The extension assumes 0% purchase at the too-cheap and too-expensive prices. Sawtooth considers 0% at the too-cheap price unrealistic and notes it pushes the suggested price upward. Treat those factors as assumptions to state in your write-up, not measured facts about your market.

Limits for B2B pricing

Van Westendorp was built around consumer price perception, and several of its weak points get worse in B2B:

The respondent may not own the budget. A facilities manager can tell you what feels expensive, but procurement or a regional director may approve the spend. See how to interview users, buyers and champions.
Samples are small. B2B segments are often a few hundred reachable people. Each answer moves the curves visibly, so report the sample size next to every chart.
The price unit is doing a lot of work. Per site, per user and per work order produce different mental comparisons. Test the unit in interviews before you survey on a number.
It measures perception, not choice. Sawtooth recommends it mainly for new products without established prices; for established categories it prefers monadic price tests or conjoint analysis, which ask people to choose between offers.

What to do with the range

Use the range to narrow what you test next. If $60 to $90 looks plausible, pick two prices in that band and test them against real buying steps: a proposal, an introduction to the approver or a paid pilot. How to measure willingness to pay compares Van Westendorp with the other methods, and pricing research shows where it fits in a full study.

A useful habit is to ask the four questions in a short live interview rather than only in a form, then ask "why that number?" after each one. The reasons, such as "that's what we pay per site for our current tool," are often more useful than the curves. If you need facilities or procurement people in facilities management outside your network, search on Instant Expert: it finds people whose work matches your question, you review them, it sends your invitations, and you pay for each call that gets booked.