Willingness to pay: survey questions that predict real spend

"How much would you pay for this?" is the most misleading question in idea validation. People answer with a number that costs them nothing to say, and founders build pricing models on top of it. To measure willingness to pay properly, anchor every question on something real: a past purchase, an existing budget, a concrete trade-off.

Here's how to do that in a survey.

Start with what they already pay

Before you ask any pricing question about your product, establish the baseline:

• "What do you currently spend to handle [problem], counting subscriptions, contractors, and your own time?"
• "What's the most you've paid for a tool in this category?"
• "Is there a budget line this would come out of, or would it be a new expense?"
• "When you last bought a tool like this, who approved it?"

A respondent who has never paid for anything adjacent is telling you something important, whatever they say later. Stated willingness to pay with no purchase history behind it is the weakest signal in customer research.

Van Westendorp price sensitivity

These four questions map the acceptable price range without ever asking "what would you pay":

• "At what monthly price would this be too expensive to consider?"
• "At what price would it be expensive, but you'd still think about it?"
• "At what price is it a good deal?"
• "At what price is it so cheap you'd doubt the quality?"

Plot the responses. The range between the "getting expensive" and "too expensive" crossover points is your viable band. It's directional and imprecise, and it still beats a single open-ended guess by a wide margin.

Gabor-Granger for a specific number

If you already have a candidate price and want to test it, ask about purchase intent at several price points:

• "How likely are you to subscribe at $19/month?" (5-point scale)
• Then repeat at $29, $39, $49.

Plot the "likely" and "very likely" responses against price and you get a rough demand curve and a revenue-maximizing point. Discount the absolute numbers, because intent overstates action, but the shape of the curve tells you something.

The commitment questions

Stated intent is inflated. You can partly correct for that by adding friction:

• "Would you join a paid pilot at [price] starting this month?" (yes / no / need more info)
• "Can I follow up with a checkout link?" (a yes here is worth ten survey yeses)
• "Would you pre-pay three months for early access at a discount?"

Any question where "yes" commits them to doing something soon gets you closer to real willingness to pay than a hypothetical scale can.

Segment the willingness-to-pay data

Cross the pricing answers with the screener and problem questions. Willingness to pay from people who deal with the problem weekly and already pay for a workaround is real. The same number from people who see it as a minor annoyance is aspirational, and a single blended average hides the difference.

Common questions

Is van Westendorp reliable for early-stage products?

It's reliable for finding a plausible range and catching pricing that's wildly off. Don't use the exact crossover numbers as your final price; use them as the boundaries of where to experiment.

How many responses do I need for pricing questions?

More than for problem validation. Aim for 40 to 60 qualified responses so a few outliers don't dominate the price distribution.

Should I show pricing before or after the value proposition?

After, because people need to understand what they'd be paying for. Keep the description to one sentence so you're testing the price and not your copywriting.

What if willingness to pay is high but nobody converts later?

That's the usual gap between what people say and what they do, and it's why the commitment questions matter. Give a "send me a checkout link" far more weight than a "7 out of 7 likely."

Test pricing as a hypothesis

Willingness to pay is one of the five hypotheses Hypothis builds and scores. It's anchored on past spend, tested with real respondents, and reported next to the problem signal so the two don't get blended together. It's free during early access. See how it works, or read idea validation survey questions for the full survey structure.