What is Willingness to Pay?

Willingness to pay (WTP) is the maximum amount a customer would give up to obtain a product or service: the ceiling on price for that person, above which they walk away. It is an economic construct and a latent one: nobody can observe another person's reservation price directly, and people are unreliable reporters of their own, so every measurement is an approximation with a characteristic bias. The distinction that organizes the field is between stated preference (what people say in a survey) and revealed preference (what they do when money is real). Stated methods are cheap, scalable, and inflated by hypothetical bias, the well-documented tendency to overstate willingness when nothing is at stake; revealed methods are expensive, narrow, and true.

The Methods

Direct questions. "What would you pay for this?" The weakest method: unanchored, strategic (respondents lowball to influence pricing, or flatter to please), and hypothetical. Useful only as a rough range and never as a price.

Van Westendorp price sensitivity meter. Four questions per respondent: at what price would the product be so cheap you'd doubt its quality, a bargain, getting expensive, and too expensive to consider. Plotting the cumulative distributions of the four answers yields an acceptable price range and indicative points within it. It is fast, widely used, and still stated preference: better at bounding a range than at fixing a price, and sensitive to how the product is described.

Gabor-Granger. Present a sequence of price points and ask purchase likelihood at each, building a demand curve and a revenue-maximizing estimate. Cleaner than open questions, still hypothetical, and prone to anchoring on the first price shown.

Conjoint and choice-based trade-offs. Respondents choose among product configurations that vary features and price together, and the analysis infers the implicit value of each attribute, including price. The strongest stated method because it forces trade-offs rather than inviting wishes, and the standard tool for pricing packaged tiers.

Revealed preference: experiments and transactions. Randomized price tests (different prices to comparable prospects, an A/B test with ethics and fairness constraints), pre-order and deposit tests, and analysis of actual purchase behavior across existing price variation. The only methods where the number involves money leaving an account, and therefore the only ones that measure WTP rather than talk about it.

Getting Better Stated Data

1. Make the product concrete.
WTP for a description is WTP for the respondent's imagination; show the thing. Pricing questions asked after participants have used a prototype and reacted to it on video (a study design a Ballpark session supports by placing tasks before the pricing block) anchor the answer in experience rather than in a paragraph.

2. Sample the buyers, not the users.
In B2B the person who uses the product often isn't the one who pays; WTP from the wrong seat measures the wrong ceiling. Audience definition matters more here than almost anywhere.

3. Randomize anchors and orders.
The first price shown moves every later answer; rotate starting points and price orders across respondents.

4. Segment the result.
WTP is a distribution, not a number; the useful output is how it differs by segment and what share of the market sits above each candidate price.

5. Discount stated numbers, explicitly.
Report stated WTP as an upper bound and validate the candidate price with something revealed before committing.

The Short Version

Willingness to pay is a real ceiling measured through unreliable windows: direct questions (barely), price-sensitivity meters (a range), Gabor-Granger (a curve), conjoint (trade-offs), and experiments (the truth). Show the product, ask the buyer, rotate the anchors, report distributions, and treat every stated figure as an optimistic bound to be tested with money. Pricing research ends, properly, at the moment someone pays.