How to Run a Pricing Survey: The Four Questions That Work

Author: Inputly

What is a pricing survey?

A pricing survey asks customers or target buyers what they would pay, and why. Done well, it returns a range with evidence behind it rather than a number someone guessed in a meeting. The method that has survived longest is the Van Westendorp price sensitivity meter: four open questions that locate where your product feels too cheap, fairly priced, expensive, and out of reach. Ask it as a single "would you pay $49 per month?" question instead, and you get polite agreement and nothing you can use.

Why do direct pricing questions fail?

Asking "what would you pay?" invites a negotiation rather than a measurement. Buyers anchor low to protect their budget, and the number they say is a position. Rating scales drift with mood and with whichever anchor appears first on the page. Selection bias also bites: people who answer from your customer list already accepted your price. None of this makes surveys pointless. It means you ask about thresholds and value, not opinions on a number you already chose. Read about survey bias before you write a single question.

TL;DR: Ask what people would pay and they give you a position, not a number. Use the four Van Westendorp questions to map where your product feels too cheap, fair, expensive and unaffordable, recruit 100+ buyers, and read a range per segment instead of one magic price.

What are the four Van Westendorp questions?

| Question | What it locates | | --- | --- | | At what price would it be so cheap you would doubt the quality? | Bottom of the believable range | | At what price is it a bargain, clearly good value? | Point of marginal cheapness | | At what price does it start to feel expensive, but still worth it? | Point of marginal expensiveness | | At what price is it too expensive to consider? | Where you lose the deal |

Ask all four about one clearly defined package, in that order, with no price shown on the page. Then add one open question: what would make this price obviously worth it? That free text is where the roadmap material hides, because people describe the integration or limit they would need to feel good about paying.

How do you set up the survey so the data holds?

1. Recruit the segment you intend to price. Recent buyers, lost deals and target accounts belong in the sample. Fans alone give you a fantasy number.
2. Describe the package concretely: what is included, billing period, seat or usage limit.
3. Keep it to six or eight questions. One package, four thresholds, one value question, one segment question.
4. Remove every anchor: no current price, no competitor prices, no suggested ranges.
5. State anonymity, then report back what you did with the answers.

Aim for 100 completed responses per segment, then split by plan, company size and use case. An enterprise buyer and a solo founder do not share a ceiling.

How do you read the results?

Plot the four cumulative curves and read the intersections. Where "too cheap" crosses "expensive" sits the bottom of the acceptable band, where "bargain" meets "too expensive" sits the top. That span is the range you can defend. Pick a point inside it on strategy rather than on the tallest bar, and treat outliers as information: a cluster of very low answers usually means the package description did not land.

How do you combine pricing data with other feedback?

A survey shows the shape of a range. It cannot tell you why you win or lose. Pair the curves with a win/loss analysis to see which price objections killed deals, and with the Kano model to check whether the features you bundled are the ones people pay extra for. Search churn answers for the word "expensive" and see what sits next to it. Then check product market fit: if users would be lost without your product, you can hold price.

What mistakes distort pricing data most?

• Surveying only happy customers and calling the result the market.
• Showing a suggested range, then reporting the answers as independent.
• Treating one number as final and skipping the interviews that explain it.
• Changing the price without saying what changed, then wondering why trust drops.

Publish the range with its sample size attached and revisit it whenever packaging changes.

Frequently Asked Questions

1. Can I run this without a research team? Yes. Six questions, one segment, 100 responses and a spreadsheet get you a usable range.
2. Should I ask "would you buy at $X"? Only after the four threshold questions, as a cross-check. Asked first, it anchors every answer that follows.
3. What if the range is below my current price? That is a finding, not a failure. Check whether the package description matched what customers receive, and look for value you deliver but do not communicate.
4. Do discounts belong in the analysis? Yes, as a separate cut. A segment that only converts on discount has its own willingness to pay.
5. Where do the answers go after the deck? Into the same feedback pipeline, tagged and linked to the requests they justify. Pricing complaints are feature requests in disguise.

Stop guessing what your product is worth. Open a free Inputly board and price from evidence instead of instinct.