ORDEEN · Choice Modelling · 3 MIN READ
What is Willingness-to-Pay (WTP)? Price Sensitivity in Choice Modelling
Ordeen
Willingness-to-pay is the monetary equivalent of a feature’s utility — the price premium a feature commands in your market.
Book a 20-min demo
Willingness-to-pay (WTP) is the maximum amount of money a customer is willing to exchange for a product, or for a specific feature upgrade within a product. In choice modelling it is the monetary equivalent of the utility a feature provides.
Ordeen runs willingness-to-pay studies with multi-chain convergence checks (r̂, ESS, divergences) and a 95% credible interval on every WTP figure, refused outright when the price slope changes sign across posterior draws. See how it fits in the full platform in our methodology.
The problem with direct questioning
You cannot reliably ask a customer “What would you pay for this feature?” Direct questioning incentivises lowballing, while Van Westendorp price sensitivity meters often lack the competitive context required to anchor realistic market decisions. Conjoint analysis solves this by forcing respondents to trade off features against price in realistic competitive scenarios, revealing their WTP through their choices. Ordeen also runs Van Westendorp and Gabor-Granger studies, which give a quicker read of stated price resistance for a single product; use conjoint when the competitive context matters.
The mathematics of WTP
In a discrete choice model, WTP is calculated as the ratio of the utility gained by adding a feature (or switching to a better level) to the marginal utility of price:
$$\mathrm{WTP} = \frac{\beta_{\text{target}} - \beta_{\text{ref}}}{|\beta_{\text{price}}|}$$
where $\beta_{\text{target}}$ is the utility of the feature you are testing, $\beta_{\text{ref}}$ is the utility of the baseline (or absent) feature, and $\beta_{\text{price}}$ is the estimated price coefficient. Because the price coefficient represents the disutility of spending money, dividing the feature’s utility by the price coefficient scales the abstract “util” into actual pounds or dollars.
Why point estimates are dangerous
Because the price coefficient sits in the denominator, WTP estimates are highly sensitive to statistical uncertainty. If the price coefficient is small, noisy, or straddles zero, the resulting WTP ratio explodes toward infinity. Reporting a single point estimate for WTP without a confidence interval is analytically reckless.
Ordeen uses the Krinsky-Robb method in its Bayesian form: it computes WTP at every posterior draw and reports a 95% credible interval for the whole sample. If the price effect changes sign across draws, the WTP is flagged as unreliable and not shown. Segment WTP is a point estimate. With conditional pricing, WTP uses one reference price: the median of all tier prices.
Market-Calibrated WTP
Traditional WTP (the ratio of coefficients) tells you what a feature is worth in isolation. Market-calibrated WTP tells you how much you can actually raise your price without losing market share to competitors. This is calculated dynamically by introducing the feature in a market simulator and searching for the exact price increase that returns the product’s share-of-preference to its original baseline. The resulting number is the price lever a product manager can actually pull, not a theoretical ceiling.
Related terms
Krinsky-Robb method · Conjoint analysis · Hierarchical Bayes · Full methodology
Willingness-to-pay (WTP) is the maximum amount of money a customer is willing to exchange for a product, or for a specific feature upgrade within a product. In choice modelling it is the monetary equivalent of the utility a feature provides.
Ordeen runs willingness-to-pay studies with multi-chain convergence checks (r̂, ESS, divergences) and a 95% credible interval on every WTP figure, refused outright when the price slope changes sign across posterior draws. See how it fits in the full platform in our methodology.
The problem with direct questioning
You cannot reliably ask a customer “What would you pay for this feature?” Direct questioning incentivises lowballing, while Van Westendorp price sensitivity meters often lack the competitive context required to anchor realistic market decisions. Conjoint analysis solves this by forcing respondents to trade off features against price in realistic competitive scenarios, revealing their WTP through their choices. Ordeen also runs Van Westendorp and Gabor-Granger studies, which give a quicker read of stated price resistance for a single product; use conjoint when the competitive context matters.
The mathematics of WTP
In a discrete choice model, WTP is calculated as the ratio of the utility gained by adding a feature (or switching to a better level) to the marginal utility of price:
$$\mathrm{WTP} = \frac{\beta_{\text{target}} - \beta_{\text{ref}}}{|\beta_{\text{price}}|}$$
where $\beta_{\text{target}}$ is the utility of the feature you are testing, $\beta_{\text{ref}}$ is the utility of the baseline (or absent) feature, and $\beta_{\text{price}}$ is the estimated price coefficient. Because the price coefficient represents the disutility of spending money, dividing the feature’s utility by the price coefficient scales the abstract “util” into actual pounds or dollars.
Why point estimates are dangerous
Because the price coefficient sits in the denominator, WTP estimates are highly sensitive to statistical uncertainty. If the price coefficient is small, noisy, or straddles zero, the resulting WTP ratio explodes toward infinity. Reporting a single point estimate for WTP without a confidence interval is analytically reckless.
Ordeen uses the Krinsky-Robb method in its Bayesian form: it computes WTP at every posterior draw and reports a 95% credible interval for the whole sample. If the price effect changes sign across draws, the WTP is flagged as unreliable and not shown. Segment WTP is a point estimate. With conditional pricing, WTP uses one reference price: the median of all tier prices.
Market-Calibrated WTP
Traditional WTP (the ratio of coefficients) tells you what a feature is worth in isolation. Market-calibrated WTP tells you how much you can actually raise your price without losing market share to competitors. This is calculated dynamically by introducing the feature in a market simulator and searching for the exact price increase that returns the product’s share-of-preference to its original baseline. The resulting number is the price lever a product manager can actually pull, not a theoretical ceiling.
Related terms
Krinsky-Robb method · Conjoint analysis · Hierarchical Bayes · Full methodology