Willingness to pay (WTP) in conjoint analysis measures the monetary value a buyer places on a feature change by comparing the utility gained from the feature to the disutility of paying a higher price. When products are sold in tiers or pack sizes (such as Starter, Pro and Enterprise software plans), standard WTP formulas break down unless the design, the reference level and the buyer base are anchored to a specific product tier.
Ordeen calculates willingness to pay across two complementary lenses — a buyer-valuation lens anchored to your focal product's current features and price, and a competitive share-holding lens inside the market simulator — with 95% credible intervals on every estimate and an automatic refusal check when the price slope is too weak to identify. See how it fits into the full platform in our methodology.
Three ways the textbook WTP formula misleads pricing teams
The textbook formula divides the utility difference between two attribute levels by the magnitude of the price slope:
$$\mathrm{WTP} = \frac{\beta_{\text{target}} - \beta_{\text{ref}}}{\lvert \beta_{\text{price}} \rvert}$$
On a single-tier product with a linear price term and a clean price slope, that ratio is a reasonable summary. On real commercial studies, three things regularly go wrong.
1. Comparing across tiers instead of within a tier
Suppose a SaaS company sells a Starter plan around £29 and an Enterprise plan around £349 using conditional pricing, where price enters on a logarithmic scale within each tier. In a logarithmic price model, a given utility gain translates into a percentage change in price, not a fixed pound amount. A feature upgrade worth +15% of plan price corresponds to £4.35 on the Starter plan and £52.35 on the Enterprise plan. Reporting a single unanchored pound figure without stating which tier's price base it was evaluated against conflates feature value with tier movement.
2. Averaging individual ratios across a heavy-tailed price distribution
Different respondents have different price sensitivities. If you divide each respondent's utility gain by their individual price coefficient and average those ratios across the sample, a handful of price-insensitive respondents (whose β_price sits close to zero) will produce huge positive or negative ratios that pull the sample average sharply upward. Mathematically, by Jensen's inequality, the average of a ratio is not the ratio of the averages.
3. Measuring value among non-buyers of the focal product
A whole-market WTP average gives equal weight to loyal buyers of your premium tier and price-sensitive buyers who would never consider your brand. If you are deciding whether to add a feature to your Pro tier, you need to know what that feature is worth to the shoppers who actually consider the Pro tier, alongside the whole-market view.
How Ordeen computes honest, tier-aware WTP
Ordeen separates willingness to pay into explicit questions so the number on screen matches the commercial decision:
Anchored to the focal product's current spec and tier price. When you select a focal product in the simulator, every attribute's reference level is set to the level that product currently carries, and log-price conversions are anchored to that product's actual price (
P₀). If a comparison would implicitly move the product across conditional-pricing tiers, Ordeen surfaces a warning rather than treating a tier jump as a simple feature swap.Focal Buyers vs. Overall Market scope. Under the Buyer Valuation lens, you can read WTP either across All Respondents or weighted by each respondent's choice probability of picking the focal product. Within each posterior draw, Ordeen aggregates the choice-probability-weighted utility gain and price slope across respondents before converting to currency, avoiding upward ratio bias.
Competitive Share-Holding WTP. Alongside utility-ratio WTP, Ordeen computes the exact price increase a focal product can take after a feature upgrade while holding its simulated market share constant against named competitors and the walk-away (
None) option.Posterior 95% credible intervals, with refusal on unidentified slopes. Every WTP figure is evaluated across posterior draws (the Bayesian form of the Krinsky-Robb method) using smooth Harrell-Davis quantiles. If the price slope is too weak to identify — specifically, if the credible interval width blows past a product-price-anchored threshold — Ordeen flags the WTP as undefined rather than printing a meaningless whisker.
Summary checklist for pricing teams
Always ask “vs. what reference level?” A £15 WTP for “24-hour battery” means nothing unless you know whether the baseline was 8 hours or 18 hours. Ordeen prints the active reference level on every WTP table header.
Check whether your price model is linear or logarithmic. When price sensitivity scales with tier price, evaluate pound WTP at the specific product's price point (
P₀), not on an abstract scale.Compare Buyer Valuation with Share-Holding Price Move. A feature may be valued at £20 by your existing fans, but in a crowded competitive market you may only be able to raise price by £12 before losing share to a rival.
Related reading: What is willingness to pay? · The Krinsky-Robb method · Conditional pricing in conjoint · Full methodology
Ready to run a pricing study with defensible WTP intervals? Book a demo or email im@ordeen.com.
