Problem
A team launching a monetized product must pick a pricing model, a pricing metric and a price point. There is no purchase history, and buyers cannot say which price structure is right for the category. A badly chosen metric or a price detached from value shows up fast in customer complaints. A price chosen on survey answers alone can differ from what people really pay, because stated willingness is intent, not behavior.
Example from the article: Intercom priced its AI agent Fin with a value-based model where the customer pays only when Fin delivers an outcome, meaning it resolves a customer query with no further help needed.
Solution in brief
Run four steps in order: qualitative research to pick the model and metric, a quantitative willingness-to-pay (WTP) survey built on that model and metric, a model that adds business realities to the survey curve, and an executive sign-off. The qualitative step comes first so the survey asks the right question. The model bridges survey intent and real revenue.
Questions this case answers
- How to set the price structure and price level of a new product before any purchase data exists, when stated willingness to pay reflects intent and not behavior
- How to choose a pricing model and a pricing metric for a new AI product
- How to find a price point from willingness-to-pay surveys
- How to turn survey-based willingness to pay into an ARR and adoption forecast
Tools and methods
- Buyer interviews — find how buyers define value and ROI and which pricing models they expect
- Gabor-Granger and Van Westendorp surveys — measure WTP at a set of price points
- Demand curve with revenue-maximizing price — adoption share times price shows where revenue peaks
- ARR and attach-rate scenario model — convert list price, discounts, usage and WTP into logos, ARR and margin
Google or a one-time email code. No password.