Willingness to Pay: How to Measure It (7 Methods)

The short answer

Willingness to pay (WTP) is the highest price a customer will pay for a product before they walk away. It varies by customer and segment. You measure it with customer interviews, pricing surveys such as Van Westendorp and Gabor-Granger, conjoint analysis, live price tests on new customers, and behavior data such as upgrade rates and discount requests. Use at least two methods and compare.

Key takeaways

Willingness to pay is a range across customers, not one number. Price for the segment, not the average.
Never ask 'what would you pay?' on its own. People anchor low and answer hypothetically.
Van Westendorp gives an acceptable price range. Gabor-Granger gives a demand and revenue curve. Use both.
Live tests on new customers are the most reliable evidence. Planet Fitness price-tested in several markets before its first rise in 26 years.
Model profit, not just revenue. In the worked example, $29 and $39 earn almost the same profit from very different customer numbers.

On 28 July 2026, Canva raised the US price of its Pro plan from $144 to $180 a year.

On 11 September, it went back to $144.

Three days later, it cut the free plan's AI allowance from 200 uses to 20. (The Canva pricing teardown has the full timeline.)

That's a willingness-to-pay test run in public, over six weeks, with its customers watching. Canva is a huge company and it will be fine. (It reported US$4 billion in ARR by the end of 2025.) But most of us can't afford to learn what customers will pay by announcing a price, reading the reaction and reversing it.

The good news is you don't have to. You can measure willingness to pay quietly, cheaply and before you change anything. Let me show you how.

What is willingness to pay?

Willingness to pay (WTP) is the highest price a customer will pay for something before they decide it isn't worth it.

Three things make it tricky:

It's different for every customer. A sole trader and a 50-person firm value the same software very differently.
What people say isn't what they do. Stated willingness to pay (survey answers) and revealed willingness to pay (what people actually buy) often differ.
It moves. It rises as customers depend on you and falls when a competitor shows up cheaper.

So don't look for "the number." Look for the range, by segment.

Why does willingness to pay matter so much?

Because your price has to sit somewhere between two lines: what it costs you to serve a customer, and what that customer is willing to pay. Everything between those lines is a choice.

Price too close to the floor and you starve the business. Price above the ceiling and nobody buys. If you've never measured it, assume you're closer to the floor than you think.

Segment pricing is how the big players capture different levels of willingness to pay without cutting their main price. Spotify charges $12.99 for Individual and $6.99 for Students in the US. Amazon Prime is $14.99 a month, with Prime for Young Adults at $7.49 and Prime Access at $6.99 for eligible customers. Same product, different willingness to pay, different price.

How do you measure willingness to pay?

There are seven practical methods. No single one is enough, so I recommend combining at least two: one that tells you why (interviews) and one that tells you how much (a survey or a live test).

Method	What it tells you	Effort	Best for	Main weakness
Customer interviews	Why people value it, what they compare it with	Low cost, high time	Early products, B2B	Small samples; people are polite
Direct question ("What would you pay?")	A rough first number	Very low	Almost nothing on its own	People anchor low and guess
Van Westendorp survey	An acceptable price range	Low to medium	New products, new tiers	Doesn't predict demand directly
Gabor-Granger survey	Share who'd buy at each price; revenue curve	Low to medium	Choosing between price points	Shows one price at a time, out of context
Conjoint analysis	Trade-offs between features and price	High	Packaging decisions, big launches	Needs specialist design and larger samples
Live price tests	What people actually pay	Medium	Existing businesses with steady sign-ups	Takes time; needs care with fairness and law
Behavior data	Upgrade rates, discount requests, objections	Low (if you track it)	Any business with customers	Backward-looking

Let's go through the ones that matter most.

Customer interviews

Start here. Ten good conversations will teach you things no survey can, because you hear the reasons.

Ask about the problem, not your product. What does it cost them today? What have they tried? What did they pay for that? Who signs off on spending? There's a ready-to-use willingness-to-pay interview script you can adapt.

The best question I know: "What would have to be true for this to be worth $X a month to you?" Then stay quiet.

Van Westendorp

The Van Westendorp Price Sensitivity Meter asks four questions: at what price would this be too cheap to trust, a bargain, getting expensive, and too expensive to consider? Plot the answers and you get an acceptable price range.

It's quick, cheap and great for new products. It doesn't tell you how many people will buy at each price, which is why I pair it with Gabor-Granger. Full guide: the Van Westendorp price sensitivity meter.

Gabor-Granger

Developed by André Gabor and Clive Granger in the 1960s, this method shows each respondent a price and asks if they'd buy. If yes, it asks about a higher price. If no, a lower one. The result is the share of people who'd buy at each price, which gives you a demand curve and a revenue curve.

Here's how that looks with real numbers.

Reading a Gabor-Granger result

Say 200 prospects answer a Gabor-Granger survey for a monthly online service. This is the share who'd "definitely" or "probably" buy at each price (illustrative):

Price a month	Would buy	Revenue per 100 prospects	Profit per 100 prospects (at $8 cost per customer)
$19	70%	70 × $19 = $1,330	70 × $11 = $770
$29	52%	52 × $29 = $1,508	52 × $21 = $1,092
$39	35%	35 × $39 = $1,365	35 × $31 = $1,085
$49	20%	20 × $49 = $980	20 × $41 = $820

Revenue peaks at $29. But look at profit: $29 and $39 earn almost the same ($1,092 vs $1,085), and $39 does it with a third fewer customers to serve and support.

So which do you choose? If growth, referrals and network effects matter most, $29. If you're capacity-constrained or support-heavy, $39. The data narrows the choice. Your strategy makes it.

One caution: survey answers can overstate real buying. Treat these percentages as relative, not as a forecast.

Turn survey answers into a price range

Paste your Van Westendorp responses and get your acceptable price range and chart, free.

Open the free tool

Live price tests

Nothing beats watching real buyers. You can:

Test on new customers only. Show a new price to new sign-ups for a month and compare conversion.
Test by location or channel. Planet Fitness price-tested in multiple markets before raising its Classic membership from $10 to $15 in 2024, its first rise in 26 years. See how Planet Fitness did it.
Test a new tier. Add a premium tier and see who takes it.

Be fair and transparent. Don't show different prices to people in ways that would feel deceptive if they found out. In Australia, misleading pricing can breach the Australian Consumer Law, so keep tests honest and get advice if you're unsure.

Behavior data you already have

You probably have more evidence than you think:

How often do prospects ask for a discount?
What share of customers pick your top tier?
Do people mention price when they cancel?
Has anyone ever said "that's cheap"?

If almost nobody pushes back on price and many choose the top tier, your willingness-to-pay ceiling is higher than your price, and it's time to read how to raise prices. The Customer Pain-Point Analyser can pull price objections out of reviews, emails and support tickets.

What questions should a willingness-to-pay survey ask?

Keep it short. A good survey has:

Two or three screening questions so you know which segment each person is in.
A clear description of the offer, including what's included and how billing works.
The pricing questions (Van Westendorp's four, or a Gabor-Granger ladder).
One purchase-intent question at your likely price.
One open question: "What would make this worth more to you?"

The willingness-to-pay survey template has the wording ready to copy.

How does willingness to pay work for AI features?

This is the live question of 2026: will customers pay extra for AI, or do they expect it included?

The market has split three ways:

Bundle it and raise the plan. Notion retired its $10 AI add-on and bundled AI into its $20 Business plan. Slack and Xero did similar things.
Give basic AI away, charge for more. Zoom includes AI Companion on paid plans and sells a $12 Custom AI Companion add-on.
Charge for one specific job. MYOB bundles general AI but sells an AI BAS add-on for A$25 a month. One clear job, one clear price.

Which works for you depends on willingness to pay for the specific outcome, not for "AI." Customers will pay for "your BAS done" far more readily than for "AI features." Test the job, not the technology.

And be careful when you use AI to justify a price rise. When Microsoft raised Australian Microsoft 365 Personal from A$109 to A$159 with Copilot added, the ACCC sued, alleging subscribers weren't clearly told about a cheaper Classic plan without Copilot. The case was before the court as at October 2026 (case study). Willingness to pay for the AI was exactly the question customers weren't given the chance to answer.

What are the most common willingness-to-pay mistakes?

Asking "what would you pay?" and stopping there. You'll get a low, made-up number.
Asking only happy customers. They already accepted your price.
Averaging across segments. The average customer doesn't exist.
Forgetting the alternative. WTP depends on what else they could do, including doing nothing.
Ignoring cost. The revenue-maximizing price isn't always the profit-maximizing one, as the example above shows.
Treating survey answers as a forecast. They're a direction, not a promise.

When is formal willingness-to-pay research overkill?

If you have a small number of high-value clients (an agency, a consultancy, an accounting practice), surveys won't help much. You don't have enough people. Instead, quote your next ten proposals at a higher price and track how many you win. Your close rate is your willingness-to-pay data.

The value-based pricing guide covers how to set those prices from the value you create.

What to do next

Pick your two or three most important segments.
Book five interviews per segment this month.
Run a short survey using Van Westendorp's four questions, then analyze it with the Van Westendorp Calculator.
If you have steady sign-ups, test your preferred price on new customers for four weeks.
Compare all three. Where they agree, you have your answer.

Before any of that, you can sanity-check your offer and wording on AI-simulated customer personas with the Synthetic Customer Panel. Just don't mistake it for the real thing.

Then ask yourself: when did a customer last tell me my price was too low? If the answer is "never," that's your willingness-to-pay data talking.

Frequently asked questions

What is the difference between willingness to pay and price?

Price is what you charge. Willingness to pay is the most a customer would accept before saying no. The gap between the two is the customer's perceived bargain. If your price sits well below most customers' willingness to pay, you're leaving money on the table. If it sits above, you'll lose sales or face churn.

How many people do I need for a willingness-to-pay survey?

Enough to see a stable pattern in each segment you care about. A common rule of thumb is around 100 completed responses per segment for survey methods like Van Westendorp. With fewer, treat the results as a direction, and back them up with interviews and a live test on new customers.

Can I measure willingness to pay with existing customers?

Yes, but be careful. Existing customers already accepted your current price, so they skew toward people who find it reasonable. They're great for testing new features, tiers or add-ons. For a new price level, include prospects and people who chose a competitor, or run a test on new sign-ups.

Does willingness to pay change over time?

Yes. It moves with the value you deliver, competitors' prices, the customer's budget and switching costs. Once customers depend on your product and have their data in it, willingness to pay usually rises. That's one reason businesses like Xero can raise prices each year with little churn. Re-measure at least annually.

Is willingness to pay the same as value-based pricing?

They're related. Value-based pricing sets prices from the value a customer receives. Willingness to pay is what they'll actually pay, which is often lower than the value because of budgets, alternatives and risk. You use value to set the ceiling and willingness-to-pay research to find where buyers really say yes.

Can AI-simulated customers replace real willingness-to-pay research?

No, but they can help you prepare. Synthetic panels are useful for spotting confusing offers, testing wording and narrowing options before you survey real people. They can't tell you what real buyers will pay. Treat their output as a hypothesis, then confirm it with interviews, surveys or a live price test.

Get both books free

The Subscription Playbook (ebook + audiobook) and Feed a Starving Crowd, instantly. In Australia? Claim the free hardcover.

Robert Coorey, MBA

Robert is the author of The Subscription Playbook and Feed a Starving Crowd and a Director at BPAG Accountants in Sydney. More about Robert.

Sources

Facts and prices checked 27 October 2026. Spotted something out of date? Tell us.

Willingness to Pay: What It Is and How to Measure It Properly