Home/Tools/Van Westendorp
Pricing

Van Westendorp — what price range will the market actually accept?

The classic four-question pricing study. Respondents are asked at what price the product would be too cheap, cheap, expensive and too expensive — Crowdmines computes the four standard intersection points (PMC, OPP, IPP, PME) and produces the familiar PSM curve chart.

PMC · OPP · IPP · PMESubgroup variableAI-assisted column mappingCumulative curves
PSM Curve · n=400Live result
SaaS launch — pricing for a new tier
Four cumulative price-sensitivity curves across surveyed prices
PMC
$24
OPP
$42
IPP
$48
PME
$71

The business question it answers

“What price range will the market actually accept, and where in that range should we land?”

Say you’re launching a new SaaS tier and need to choose between $29, $49 and $99 / month. Run PSM on a 400-response study and the chat returns PMC ≈ $24, OPP ≈ $42, IPP ≈ $48, PME ≈ $71. The acceptable range sits between PMC and PME; the optimal price is at OPP. You walk into the pricing committee with a defended price point and a “do-not-cross” ceiling.

Set a subgroup variable in the chat — “split by region” — and the same analysis runs once per group with a side-by-side comparison view. Useful when you suspect price sensitivity differs by segment.

How the methodology works

Each respondent answers four price questions: too cheap, cheap, expensive, too expensive. Crowdmines enforces the constraint too cheap < cheap < expensive < too expensive and drops invalid responses before computing curves.

  • PMC — Point of Marginal Cheapness. Below this, buyers worry about quality. Floor of the acceptable range.
  • IPP — Indifference Price Point. Equal numbers see the price as cheap vs. expensive. Psychological middle.
  • OPP — Optimal Price Point. Total resistance is minimised. The price with the least pushback.
  • PME — Point of Marginal Expensiveness. Above this, too many buyers walk away. Ceiling of the acceptable range.

Intersections are located by linear interpolation between adjacent price bins where two curves swap which is higher — so the result is a precise price, not just the nearest surveyed value. Below 30 valid responses, the chat asks you to widen filters rather than producing unreliable curves.

What you see in the chat

A PSM curve chart with the four intersection price points marked, a small table of the four price points, summary statistics (n, mean, median), and a one-paragraph written interpretation. Request a subgroup and you get the same artefacts per group plus a comparison summary.

Required data
ColumnsFour numeric PSM price questions
CurrencySame currency, no mixed units
Min sample~30 valid responses
OptionalCategorical subgroup variable
Subgroup examplesRegion · segment · channel · plan tier
Use when you hear
  • We’re guessing on price. / Pricing is a gut call right now.
  • We need a defensible price for the launch deck.
  • We want a price range, not just one number.
  • Different markets / segments seem to react differently to price.
  • Our finance team wants a sanity-check on the price our PM picked.
Disqualifier

Only one price question?

That’s a Gabor Granger fit, not PSM — Gabor Granger is faster and more direct when you only want a single recommended price.

See Gabor Granger →
Step by step

From question to chart in seven steps.

Here's what you experience in the chat, from start to finish.

1
Ask the question
“What’s the right price range for Product X?” — the platform recognises the intent and selects Van Westendorp automatically.
2
Map the columns
Confirm which columns map to too cheap, cheap, expensive, too expensive. The platform suggests mappings if names are close.
3
Apply filters
Optional. e.g. “Only respondents from the US” or “Only people who saw Concept B”.
4
Choose a subgroup
Optional. “Split by region” or “Compare age groups” — analysis runs independently per group.
5
Confirm and run
Final summary: column mappings, active filters, subgroup variable, expected sample size.
6
Review results
PSM curve chart, price-points table, summary statistics, written interpretation.
7
Follow up
“Now split that by gender” — the chat remembers the dataset and prior analysis.
Compared to

How Crowdmines compares to Excel, SPSS and survey-platform add-ons.

Van Westendorp PSM has been around since the 1970s. Most research teams run it one of three ways: manually in Excel, through a general-purpose stats package (SPSS, R, Python scripts), or via a survey platform that bundles basic analytics (Qualtrics, SurveyMonkey, Conjointly, Displayr).

CapabilityTraditional (Excel / SPSS / R)Survey-platform add-ons (Qualtrics, Conjointly)Crowdmines
Setup effortExport data, open template or write a script, map columns manuallyTied to one platform's data format; configure the analysis moduleAsk in the chat — column mapping is AI-assisted, no scripting
Time to result30–60 minutes; hours when debuggingMinutes once configured; setup can be slowSeconds from question to chart
Subgroup compareRe-run per group, manually combine outputsSometimes; often a separate export per segment“Split by region” → runs per group with a comparison view
Iterative refinementChange filter, re-export, re-runRe-configure, re-runType a follow-up sentence in the same conversation
IntegrationSeparate tools, separate exportsLimited to what the platform bundlesPivot from PSM → Gabor-Granger → Driver Analysis without re-uploading
Analyst dependencyRequires a trained analyst or data scientistResearcher still needs to configure correctlySelf-serve — anyone who can describe the question can run it
Beta Program Open

A defended price point and a “do-not-cross” ceiling for the pricing committee.

Ask the question in the chat — column mapping is AI-assisted, no scripting. Pivot from PSM to Gabor Granger to Driver Analysis without re-uploading.