How to Find What Customers Are Willing to Pay
You can find out what customers are willing to pay in two ways: watch what they actually buy at different prices, or ask them. Watching is more reliable. Your sales history, or a live price test, shows real purchases. Asking works when you don't have sales yet: a Van Westendorp or Gabor-Granger survey gives you a price range before launch. Just expect people to overstate what they'd pay. Across 77 studies, stated willingness to pay ran 21% higher than what people actually paid.
What Willingness to Pay Actually Means
Willingness to pay (WTP) is the highest price a particular customer would pay for a product before deciding not to buy. Every customer has a different number. What you really want to know is the shape of that spread: how many customers you keep at $30, at $35, at $40. That shape is your demand curve, and the price that makes the most profit sits somewhere on it.
Researchers split the ways of measuring it into two families. A widely cited review by Breidert, Hahsler and Reutterer separates revealed preference methods, which observe real purchases, from stated preference methods, which ask people what they would do. For an online store, that becomes five practical options.
5 Ways to Measure What Customers Will Pay
| Method | Type | Best for | Main limitation |
|---|---|---|---|
| Sales history (price elasticity) | Revealed | Existing products that have changed price before | Needs past price variation; promotions distort it |
| Live price test (A/B) | Revealed | Products with steady traffic | Needs enough orders to be conclusive; showing different prices can upset customers |
| Van Westendorp survey | Stated (direct) | New products, finding an acceptable range | People overstate; gives a range, not demand at each price |
| Gabor-Granger survey | Stated (direct) | Checking specific candidate prices | Respondents anchor on the first price shown |
| Conjoint analysis | Stated (indirect) | Comparing features and price together | Complex to design; needs larger samples |
1. Read your sales history
If a product has sold at more than one price, you already have real evidence. Compare how many units sold at each price, excluding sale periods, and you get its price elasticity: how much demand moves when price moves. It's the most trustworthy method because it's based on what customers did with their own money. The price elasticity of demand formula shows how to calculate it from two price points.
2. Run a live price test
If a product has only ever had one price, you can create the evidence: show a different price for a set period, or to a share of visitors, and compare conversion and profit. It's the gold standard when you have the traffic, but it needs enough orders to separate a real difference from noise. How to run a price A/B test covers sample sizes and how long to run it.
3. Ask with a Van Westendorp survey
The Van Westendorp Price Sensitivity Meter asks four questions: at what price would the product be so cheap you'd doubt its quality, a bargain, getting expensive, and too expensive to consider. Plotted together, the answers give an acceptable price range and a point where roughly as many people find it too cheap as too expensive. It's quick, needs no sales history, and works before launch. Our free survey template has the exact wording, how to run one covers who to ask, and how to interpret the results explains the four price points.
4. Test specific prices with Gabor-Granger
Gabor-Granger shows respondents a specific price and asks if they'd buy, then moves the price up or down depending on the answer. The result is an estimated share of buyers at each tested price, which you can multiply out to find the revenue- or profit-maximizing option. It's useful when you've narrowed it down to a few candidate prices and want to compare them directly.
5. Use conjoint analysis for features and price together
Conjoint asks people to choose between product versions that differ in several features and price at once, then works out how much each feature is worth to them. It's powerful for deciding what goes into a premium tier or a bundle, but it takes careful design and more responses than a small store usually has access to.
See what Zorin's elasticity model says about your own catalog.
Start free trialWhy Surveys Overestimate, and How to Correct for It
Saying you'd pay $40 costs nothing. Actually paying $40 does. That gap has a name, hypothetical bias, and it has been measured. A 2020 meta-analysis by Jonas Schmidt and Tammo Bijmolt, covering 77 studies and more than 20,000 real purchase decisions, found hypothetical willingness to pay was 21% higher than real willingness to pay on average. Surprisingly, indirect methods like conjoint overestimated more than direct questions.
Three practical corrections:
- Treat survey results as a ceiling, not a target. If a survey says customers find $40 acceptable, launch somewhere below that and test up.
- Survey real target customers. Existing buyers and people who fit your customer profile give far more useful answers than friends or a general panel.
- Replace the survey with sales data as soon as you can. Once real orders arrive at a couple of price points, they outrank anything the survey said.
Which Method Should You Use?
- New product, no sales yet: Van Westendorp for a range, Gabor-Granger if you're choosing between a few specific prices.
- Existing product that has changed price before: your sales history. It's free and based on real purchases.
- Existing product that has never changed price: a deliberate small price change or a live test, since there's nothing to read yet.
- Designing tiers, bundles or a premium version: conjoint, or a simpler survey comparing a few versions.
Whichever method you use, remember that willingness to pay isn't one number for your whole customer base. Buyers from a discount-heavy ad channel and loyal repeat customers can differ a lot, and a blended average can hide both.
Key Takeaways
- There are five practical ways to measure willingness to pay: sales history, live price tests, Van Westendorp surveys, Gabor-Granger surveys, and conjoint analysis.
- Methods based on real purchases (sales history, price tests) beat methods based on what people say, but they need existing products and traffic.
- A meta-analysis of 77 studies found hypothetical willingness to pay overstates real willingness to pay by 21% on average, so discount survey answers.
- For a new product with no sales, start with a survey. Once real orders arrive, let sales data take over.
- Willingness to pay varies by customer and channel, so one blended number can hide very different groups.
Frequently Asked Questions
How do I find out what customers are willing to pay?
Look at how sales changed at different prices you've charged before, or run a live price test. If the product is new, ask potential customers with a Van Westendorp or Gabor-Granger survey, then switch to real sales data once orders come in.
What is the most accurate way to measure willingness to pay?
Methods based on real purchases, such as sales history and live price tests, are the most accurate because customers are spending real money. Surveys are useful when no sales data exists, but stated willingness to pay runs about 21% higher than real willingness to pay on average.
What's the difference between Van Westendorp and Gabor-Granger?
Van Westendorp asks open-ended questions about when a price feels too cheap or too expensive and produces an acceptable range. Gabor-Granger shows specific prices and asks whether the respondent would buy, producing an estimated share of buyers at each price.
How many survey responses do I need?
For a Van Westendorp survey, a few hundred responses from genuine target customers give a usable range, and smaller samples can still point you in the right direction for a small store. What matters more than the count is asking people who actually resemble your buyers.
Can I measure willingness to pay without a survey?
Yes, if the product has sold at more than one price. Your sales history already shows how demand responded to each price, which is a direct read on willingness to pay across your customer base.
Is willingness to pay the same for every customer?
No. It varies by customer and by channel. Buyers who arrived through a discount-heavy ad campaign and loyal repeat customers can respond to price very differently, so it is worth checking segments separately rather than relying on one blended number.
What customers say they'll pay is a useful starting point. What they actually pay is the answer. Start with a survey if you have nothing else, then let real sales take over. If your products already have sales history, Zorin reads it from Shopify or WooCommerce and shows each product's price sensitivity, and its built-in Van Westendorp survey covers the products that don't.
Written by Dexter
Dexter is part of the team at Zorin, building tools that help ecommerce merchants price with data instead of guesswork.
More in Price Sensitivity Surveys
View all →Do You Need a Survey If You Have Sales Data?
A pricing survey and your sales history measure different things. See when a Van Westendorp survey beats elasticity data, and when it doesn't.
How to Run a Van Westendorp Price Survey
The 4 Van Westendorp questions, who to ask, how many responses you need, and how to turn the answers into an acceptable price range before you launch.
How to Interpret Van Westendorp (PSM) Results
Learn what the four Van Westendorp price points actually mean, how to read a narrow vs wide range, and why the optimal price point isn't your final price.