Gabor-Granger vs. Van Westendorp: Choose the Pricing Question First

Use Van Westendorp to investigate perceived price boundaries for a defined offer. Use Gabor-Granger to compare stated buying responses across specified prices. When their outputs point in different directions, check the question and commercial objective before paying for another study or averaging the answers.

Consider a team choosing the online price of a rechargeable desk light. A perception report places the light in a $20–$24 range, while a price-ladder worksheet favors $16 on its price-times-response index. Neither result cancels the other. They describe different reactions, and costs can change which candidate deserves a market test.

The desk-light offer, records and costs below are fictional teaching inputs. The two files illustrate separate instruments for the same offer; no human research or Sapien simulation was run. Five records per file keep the arithmetic inspectable and aren't a sample-size recommendation.

The same light, two research assignments

The offer is one adjustable rechargeable desk light with three brightness settings and a USB-C charging cable, without a wall adapter. It is a one-off purchase from the brand's website, with standard delivery included. Study prices are USD amounts before consistently displayed applicable sales tax. Contents, buying occasion and presentation stay fixed.

A buyer who hasn't seen proposed prices can answer four Van Westendorp questions: when the light seems suspiciously cheap, good value, expensive but still worth considering, and too expensive to consider. The answers are monetary thresholds. Conjointly's method documentation describes the four perception curves and their intersections.

A Gabor-Granger participant instead sees an amount and answers a buying question: “At $20 before tax, with delivery included, would you choose this light on your next desk-light purchase?” A planned route then shows another amount according to the answer. The output retains prices shown, answers and the accepted/rejected bracket. Conjointly's Gabor-Granger documentation describes that repeated-price structure for a fixed offer.

Those assignments aren't interchangeable. A cheapness threshold isn't a purchase answer at that amount. A last accepted ladder price doesn't tell you when a low price creates quality doubt.

What the perception file supports

These five records supply four thresholds each. The slash separates the two named fields in each column.

For this illustration, cheapness curves count thresholds at or above a grid price; expensiveness curves count thresholds at or below it. Equality is included, and the first whole-dollar equality defines the boundary.

At $20, two of five too-cheap thresholds remain at or above the price, and two expensive thresholds are at or below it. At $24, two good-value thresholds remain at or above the price, and two too-expensive thresholds are at or below it. The resulting perception interval is $20–$24 under this convention.

It doesn't contain a count of people buying at $20 or $24. Nor does an intersection called “optimal” identify the price with the highest contribution. Our Van Westendorp guide explains the curve and equality rules in a complete worked analysis.

If a supplier also presents a demand or revenue curve, ask what additional inputs and assumptions produced it. The pricesensitivitymeter implementation requires extra purchase-intent inputs for the Newton–Miller–Smith extension. The four perception amounts alone don't supply those responses. An extension also doesn't turn a survey into observed sales.

What the ladder file supports

The second teaching file uses $12, $16, $20, $24 and $28. Every route starts at $20 and stops at adjacent acceptance/rejection or the highest tested amount. Insufficient information would stop the route as unresolved; none of these five teaching routes uses that state.

The last accepted shown amounts are $12, $16, $20, $24 and $28. Count amounts at or above each candidate, assuming acceptance at a higher amount implies acceptance at lower amounts. This summary rule fills in some unshown price responses; it isn't five directly observed answers at every amount.

The price-times-share index peaks at $16, below the perception interval. The files have different records and questions, so this isn't evidence that particular people contradicted themselves. Even in a properly matched study, an aggregate perception boundary and a stated buying curve needn't select the same amount.

The index also omits costs. Assume the seller keeps the tested pretax amount, with no deductions beyond an $8 variable cost per delivered light. At $16, the contribution index is ($16 − $8) × 80% = $6.40. At $20, it is ($20 − $8) × 60% = $7.20. The other candidates yield $4.00 at $12, $6.40 at $24 and $4.00 at $28.

Under those supplied assumptions, $20 leads on contribution while $16 leads on the price-times-response index. Both are indexes from stated responses. Our Gabor-Granger guide follows the full ladder and reporting requirements; actual buying still needs a suitable commercial test.

The useful comparison is a commissioning decision

If the team has no credible price range, the perception task can reveal whether buyers' expectations even overlap the business's feasible amounts. If the team already has implementable candidates, another four-question perception exercise may leave the price choice exactly where it began. Commission a comparison of responses to those candidates instead.

For the desk light, a contribution objective makes $20 the leading ladder candidate under the $8 cost assumption. A trial or reach objective gives $16 a role. The next brief can compare those two complete offers and the costs each must cover, rather than collect both methods again because the first charts look different.

Check the response source, audience and exposure before attributing a disagreement to the method. Were both files about the same light, tax treatment and delivery? Did they cover the same intended buyers? Did people see a preferred range before naming thresholds? Did the ladder start at a higher amount for one group? Preserve those conditions and any missing answers.

Using both methods in the same questionnaire creates another exposure decision. A person who has just named price thresholds approaches the ladder with a different history. Agree the order or independent groups before collection. The two teaching files here demonstrate outputs; they aren't a combined questionnaire plan.

If the remaining question concerns rival brands, changing features or several bundles, neither isolated-price task may complete it. Our pricing-method comparison shows when a choice-based design or another route fits better. Specify the real alternatives and the decision instead of choosing a technique by familiarity.

Sapien can compare simulated reactions to complete priced offers and investigate differences across buyer groups. For a proposed study, agree which task will be run, its response source, required records and what the recommendation must resolve. A method label doesn't establish that every implementation or analysis is included.

If your team has two feasible prices and different reasons to favor each, scope the comparison with Sapien. The useful deliverable explains which price advances, which objective favors it and which actual purchase evidence remains necessary.