Van Westendorp Is a Lie Detector, Not a Pricing Tool

It asks four honest questions and gets four dishonest answers. Here's what the Price Sensitivity Meter actually measures — and why it quietly underprices premium brands.

What the Van Westendorp method actually is

The Van Westendorp Price Sensitivity Meter is a survey technique invented in 1976 by Dutch economist Peter van Westendorp. Fifty years later it is still the default tool when a brand wants a quick read on what a new product should cost. It is simple, cheap, and — used honestly — a reasonable place to start a conversation.

It works by asking every respondent four price questions about the same product:

Each answer is plotted as a cumulative curve across the full price range. Where those curves cross, you get the method's headline outputs:

• Optimal Price Point (OPP) — where the "too cheap" and "too expensive" curves intersect. The point of minimum buyer resistance.
• Indifference Price Point (IPP) — where the "bargain" and "getting expensive" curves cross. Often read as the typical market price or the median respondent's reference price.
• Range of Acceptable Prices — bounded by the Point of Marginal Cheapness (too cheap × getting expensive) and the Point of Marginal Expensiveness (bargain × too expensive). The corridor inside which most people will at least consider buying.

That's the whole method. Four questions, four curves, a price range. No model, no behavior, no purchase. Which is exactly where the trouble starts.

Where it's genuinely useful

Let's be fair before we're critical. For an early-stage product with no pricing history, a tight budget, and a team that needs a defensible starting number by Friday, Van Westendorp earns its keep. It's fast, the fieldwork is cheap, and it gives you four anchor points and a plausible corridor instead of a guess. As a hypothesis generator — a first sketch of the demand landscape — it's fine. The mistake is treating that sketch as a verdict.

The structural flaw

Van Westendorp is pure stated preference. It asks people to price a product in the abstract: no competitors on the shelf, no feature list, no real budget, no consequence for being wrong. Nobody hands over a card. The say-do gap — the chasm between what people say and what they do — hits pricing harder than any other research domain, because the moment you ask "what would you pay," you've created an incentive to lowball.

And lowball they do. When there's no real purchase on the line, respondents anchor on the cheapest defensible number. That's why the method systematically underprices premium and aspirational products. The whole point of a premium brand is that it's worth more than a respondent would admit to a stranger with a clipboard. A craftsmanship story, a status signal, a design language people pay for emotionally — none of it survives the abstraction of a survey grid. Even Google's AI Overview on this exact keyword concedes the method "ignores competitor pricing and product features." When the tool itself can't see your differentiation, it will price you like a commodity.

The three errors

The underpricing isn't one bug. It's three, stacked.

• No consequence → lowballing. A stated price costs the respondent nothing. So the rational move is to claim a low ceiling. There's no mechanism in the four questions to surface what someone would actually pay when the product is in their hands and they want it.
• No competitive context. Real prices live next to alternatives. Van Westendorp strips the shelf bare — no rivals, no substitutes, no reference brands. People price in a vacuum, and the number that comes back has never met a competitor.
• Anchoring from the question order. The four questions march from "too cheap" upward. That sequence plants reference points before the respondent has formed a view, dragging answers toward whatever range the survey implicitly suggests. The instrument shapes the very thing it claims to measure.

What to do instead

You don't have to throw Van Westendorp out. You have to stop trusting it alone — and pair or replace it with evidence about what people do, not what they say.

• Implicit measurement. Reaction-time-based testing of price-value associations captures the gut-level acceptance a survey question can't reach. It reads the unconscious "yes, that's worth it" that respondents won't articulate — and that premium brands live on.
• In-market A/B tests. Where you can technically run them — digital products, e-commerce, controlled retail — actual purchases at different price points are the gold standard. Revealed preference beats stated preference every time.
• Causal AI. Instead of reading off current stated willingness-to-pay, Causal AI models how a price change would actually shift the demand curve — the causal drivers of acceptance, not a snapshot of today's claims. That's the difference between "what would you pay" and "what happens if we move the price."

This is SUPRA's Deep Implicit + Causal approach: measure the unconscious response, then model the consequence. It's built for exactly the decisions Van Westendorp can't carry.

So should you use it?

The rule is simple. The lower the stakes, the more you can lean on Van Westendorp. The higher the stakes — and the more premium your brand — the faster you should walk past it.

Frequently asked questions