The Van Westendorp price sensitivity meter is a pricing survey created by Dutch economist Peter van Westendorp in 1976. It asks four questions: at what price is a product too cheap to trust, a bargain, getting expensive, and too expensive to consider. Plotting the answers as cumulative curves gives an acceptable price range and four reference points, including the optimal price point.
Key takeaways
- Van Westendorp asks four price questions and turns the answers into four cumulative curves.
- The acceptable price range runs from the point of marginal cheapness to the point of marginal expensiveness.
- In the worked example, 200 answers give a range of $22 to $33, with an optimal price point of $26.
- It shows what feels acceptable, not how many people will buy. Pair it with Gabor-Granger or the Newton-Miller-Smith extension.
- Describe the offer precisely, survey each segment separately, and drop answers that aren't in logical order.
Ask a group of potential customers "What would you pay for this?" and you'll get a number. Usually a low one, and usually a guess.
Now ask a different question: "At what price would this be so cheap you'd doubt the quality?"
The conversation changes. People stop bargaining and start thinking about value. A $9 accounting service sounds suspicious. A $9 coffee subscription sounds like a catch. Price tells people something about quality, and your customers know it.
That insight sits at the heart of a method a Dutch economist named Peter van Westendorp introduced in 1976. It's still one of the quickest ways to find a price range you can defend, and you can run it this week with a free survey tool and a spreadsheet.
Here's exactly how.
What is the Van Westendorp price sensitivity meter?
The Van Westendorp price sensitivity meter (PSM) is a survey method that finds the range of prices customers see as acceptable for a product. Instead of asking for one number, it asks four questions about price thresholds. You plot the answers as four curves, and where the curves cross tells you the acceptable range.
It doesn't tell you how many people will buy at each price. It tells you which prices feel too cheap, too expensive, or about right. That's a different and very useful question, especially for a new product or a new tier.
What are the four Van Westendorp questions?
After describing the product and its price basis clearly, ask:
- Too cheap: "At what price would you consider this product to be priced so low that you would feel the quality couldn't be very good?"
- Cheap (a bargain): "At what price would you consider this product to be a bargain, a great buy for the money?"
- Expensive: "At what price would you consider this product to be starting to get expensive, so that it's not out of the question, but you'd have to give some thought to buying it?"
- Too expensive: "At what price would you consider this product to be so expensive that you would not consider buying it?"
Let people type any number. Don't offer a list of prices, because that anchors them.
One tip that matters more than it sounds: describe the offer precisely. "An online bookkeeping tool for tradies, including bank feeds and GST reports, per month, billed monthly." If people imagine different products, you'll get different curves. The willingness-to-pay survey template has wording ready to copy.
How do you plot and read a Van Westendorp chart?
For each price level, calculate four percentages:
- Too cheap: the share of respondents whose "too cheap" answer is at or above this price. (So at this price, they'd think it's too cheap.)
- Cheap: the share whose "bargain" answer is at or above this price.
- Expensive: the share whose "getting expensive" answer is at or below this price.
- Too expensive: the share whose "too expensive" answer is at or below this price.
The first two curves fall as price rises. The last two climb. Where they cross gives you four reference points. These are the definitions used in the Wikipedia summary of the method:
| Point | Where the curves cross | What it means |
|---|---|---|
| Point of marginal cheapness (PMC) | Too cheap × Expensive | Lower bound of the acceptable range |
| Optimal price point (OPP) | Too cheap × Too expensive | Equal shares think it's too cheap and too expensive |
| Indifference price point (IPP) | Cheap × Expensive | Equal shares see it as a bargain and as getting expensive |
| Point of marginal expensiveness (PME) | Too expensive × Cheap | Upper bound of the acceptable range |
The range of acceptable prices runs from PMC to PME.
A note of honesty: practitioners don't all draw these the same way. Some use inverted "not cheap" and "not expensive" curves to find the bounds, which gives a wider range. Pick one convention, write it down and stick with it, so you can compare surveys over time.
Worked example: 200 answers, one price range
Let's say you're testing a monthly online membership for tradespeople and 200 people complete the four questions. Here's the summary at $5 steps (illustrative data).
Van Westendorp results for a monthly membership (200 respondents)
| Price a month | Too cheap | Cheap (bargain) | Getting expensive | Too expensive |
|---|---|---|---|---|
| $10 | 78% | 100% | 0% | 0% |
| $15 | 46% | 86% | 0% | 0% |
| $20 | 21% | 65% | 8% | 0% |
| $25 | 6% | 42% | 19% | 4% |
| $30 | 2% | 20% | 38% | 12% |
| $35 | 2% | 12% | 55% | 18% |
| $40 | 0% | 6% | 65% | 30% |
| $45 | 0% | 3% | 77% | 45% |
| $50 | 0% | 1% | 86% | 57% |
Working at $1 steps between these rows, the curves cross at:
| Point | Price |
|---|---|
| Point of marginal cheapness (too cheap × expensive) | $22 |
| Optimal price point (too cheap × too expensive) | $26 |
| Indifference price point (cheap × expensive) | $28 |
| Point of marginal expensiveness (too expensive × cheap) | $33 |
Acceptable range: $22 to $33 a month.
Using the alternative "not cheap" and "not expensive" curves, the same data gives a wider range of about $18 to $41. That's why you should state your convention.
What would I do with this? I'd shortlist $27 and $29. Both sit between the optimal and indifference points, comfortably inside the range. Then I'd check them against my cost to serve, and test them with real sign-ups or a Gabor-Granger question before deciding.
Run your Van Westendorp analysis in seconds
Paste your four columns of answers and get the curves, the four price points and your acceptable range, free.
How do you choose a price from Van Westendorp results?
The range is the answer, not a single point. Within it:
- Near the bottom (PMC to OPP) if you're chasing volume, entering a crowded market or want a low-friction entry tier.
- Around OPP to IPP for your main plan. It's the zone where the fewest people feel the price is off.
- Near the top (IPP to PME) for a premium tier, or if your brand and service justify it.
Then layer in two things the survey can't see: your cost to serve (the Subscription Price Calculator works this out after GST and fees) and your strategy. If $26 leaves you a thin margin, the range tells you that customers will still find $30 acceptable.
Van Westendorp works especially well for setting tiers. Survey each tier's description separately and you'll often find that the ranges barely overlap, which tells you your good-better-best tiers are genuinely different products.
How do you do a Van Westendorp analysis in Excel or Google Sheets?
- Put each respondent in a row with four columns: too cheap, cheap, expensive, too expensive.
- Clean the data. Drop rows where the four answers aren't in order (too cheap ≤ cheap ≤ expensive ≤ too expensive), and obvious jokes ($0, $1,000,000).
- Make a price column from your lowest to highest plausible price, in $1 steps.
- Calculate each curve. For "too cheap" at price P, use
COUNTIF(range,">="&P)/COUNT(range). Same for "cheap". For "expensive" and "too expensive", use"<="&P. - Chart all four as lines on one chart.
- Find the crossings by looking for the row where the difference between two curves changes sign.
Or skip the spreadsheet and paste the four columns into the Van Westendorp Calculator.
What is the Newton-Miller-Smith extension?
Van Westendorp tells you what's acceptable, not what sells. The Newton-Miller-Smith (NMS) extension fixes part of that. It adds two questions asking how likely the respondent is to buy at their own "bargain" and "getting expensive" prices, on a 1 to 5 scale.
NMS assumes nobody buys at their "too cheap" or "too expensive" prices. Combining the purchase likelihoods with the price answers gives a rough demand curve and a revenue curve, so you can estimate the price that brings in the most revenue.
If you want a demand curve without the extra complexity, a separate Gabor-Granger question (asking whether people would buy at a series of set prices) does a similar job. The willingness to pay guide shows a worked Gabor-Granger example.
When should you use Van Westendorp, and when shouldn't you?
| Use it when | Don't rely on it when |
|---|---|
| You're pricing a new product or tier | Buyers can't picture the product from a description |
| You need a defensible range quickly and cheaply | Price depends heavily on volume or custom scope |
| You can reach around 100 people per segment | You only have a handful of potential customers |
| Customers have a rough sense of what similar things cost | The category is brand new and nobody has a reference point |
| You'll validate with a live test afterwards | You plan to treat the output as a sales forecast |
Its biggest weakness is that people are answering hypothetically. They're not spending money. So I treat Van Westendorp as a way to narrow the field, never as the final word.
What makes a Van Westendorp survey go wrong?
- A vague offer description. Different mental products, messy curves.
- Mixing segments. A sole trader and a 50-person firm in one dataset blur both ranges.
- Asking only existing customers. They've already accepted your current price.
- Leaving the billing period unclear. Monthly and annual answers in one column are meaningless.
- Unclear tax basis. In Australia, state whether prices include GST.
- Treating the optimal price point as the answer. It's a reference point inside a range.
You can pressure-test your offer description before you survey real people with the Synthetic Customer Panel. It's good at spotting confusing wording, though it can't tell you what real buyers will pay.
What to do next
- Write a one-paragraph description of the offer, including billing period and tax basis.
- Set up the four questions (plus a few screening questions for segments).
- Collect answers from real prospects, not just current customers.
- Clean the data and run it through the Van Westendorp Calculator.
- Shortlist two prices inside the range, check them against cost, then test with real sign-ups.
For the full pricing process around this, see how to price a subscription or, for software, how to price a SaaS product. If you'd rather anchor on outcomes than opinions, read value-based pricing.
And ask yourself one question: at what price would my own product start to look too cheap to trust? If you're charging below that line today, you may be losing customers by being cheap.
Frequently asked questions
In what order should I ask the four Van Westendorp questions?
Most practitioners ask them in a logical sequence, either from too cheap up to too expensive or the reverse, so respondents build a consistent picture. Whichever order you choose, keep it the same for everyone. After collecting answers, check each respondent's four prices run in order, and drop those that don't.
Can I run Van Westendorp for a subscription price?
Yes. State the billing period clearly in every question, such as 'per month, billed monthly'. If you offer annual plans, survey the monthly price and work out the annual discount separately. Mixing monthly and annual prices in one survey confuses respondents and the curves.
Should prices in an Australian survey include GST?
For consumers, yes. They think in GST-inclusive prices, so ask for and report GST-inclusive numbers, and say so in the question. For business buyers who think ex-GST, ask ex-GST and label it. Either way, make the basis clear so the answers are comparable.
How is Van Westendorp different from conjoint analysis?
Van Westendorp asks directly about price for one defined offer. Conjoint analysis shows respondents combinations of features and prices and infers how much each feature is worth from their choices. Conjoint is better for packaging decisions but needs more design work and larger samples. Van Westendorp is faster and cheaper.
Can I use Van Westendorp for B2B products?
Yes, with care. B2B buyers often don't control the budget alone, and prices depend on volume. Define a typical package, ask the person who influences or approves the purchase, and segment by company size. With small B2B markets, back the survey up with interviews and real proposals.
What if the curves don't cross neatly?
It usually means your sample is small, mixes very different segments, or the offer description was unclear. Split the data by segment and look again. If the curves still behave oddly, rewrite the offer description and rerun the survey with a cleaner audience before trusting any point.
Get both books free
The Subscription Playbook (ebook + audiobook) and Feed a Starving Crowd, instantly. In Australia? Claim the free hardcover.
Robert Coorey, MBA
Robert is the author of The Subscription Playbook and Feed a Starving Crowd and a Director at BPAG Accountants in Sydney. More about Robert.
Sources
Facts and prices checked 27 October 2026. Spotted something out of date? Tell us.