Willingness to pay is not a permanent number hidden inside a customer. It is a decision made in context.
An acceptable price is not a property of the product. It is a property of the moment in which someone is asked to pay.
The pull comes from the problem: how urgent it is, how valuable the promised outcome would be, and how confident the buyer is that it will actually arrive. Confidence is the part vendors underweight — the same outcome is worth far less to someone who half believes it.
The drag comes from everything the purchase costs beyond the invoice: the alternatives already available, what the package does and does not include, the risk of being wrong, when payment falls due, whose budget it comes from, the effort of implementation, and the cost of leaving whatever is in place today.
Two forces sit outside the transaction entirely. Competing priorities decide whether the buyer looks at this at all, and the vendor's credibility decides how much of the promise they discount before they start comparing.
That is why asking “How much would you pay?” usually produces weak evidence. The respondent is being asked to invent a purchase without a concrete offer, organizational process or consequence.
Pricing research should reduce uncertainty about a real commercial decision. Its output is not one magical number. It is a documented view of:
- which segment values the outcome;
- how buyers evaluate alternatives;
- which pricing unit feels fair;
- what package they can approve;
- which price ranges deserve behavioral testing;
- why an offer is accepted or rejected;
- what economics the business can sustain.
What willingness to pay actually means
Use three separate concepts.
Ability to pay
The customer or organization has access to sufficient funds. A company may have ample resources but no relevant budget, or a department may value the product but lack purchasing authority.
Willingness to pay
Given a specific offer and context, the buyer prefers purchasing over keeping the money or choosing an alternative.
Readiness to pay now
The problem, timing, authority and process are aligned enough for commitment today.
A prospect can have ability and theoretical willingness but no current readiness because a contract is locked for nine months. Another can have urgent readiness but insufficient trust in an early product.
Record these states separately. “Interested” is not a commercial stage.
The evidence hierarchy
Not all pricing evidence deserves equal weight.
| Evidence | What it reveals | Main limitation | Relative strength |
|---|---|---|---|
| General positive feedback | Problem or concept resonates | Courtesy and no trade-off | Very low |
| Direct stated price | Respondent's constructed opinion | Hypothetical and anchor-sensitive | Low |
| Current alternative and spend | Existing budget and behavior | New outcome may differ | Medium |
| Reaction to concrete packages | Trade-offs and objections | Still hypothetical | Medium |
| Introduction to budget owner | Internal seriousness | No purchase yet | Medium to high |
| Accepted proposal or pilot scope | Commercial fit | Can still stall | High |
| Deposit or paid pilot | Real commitment | Small sample and pilot effects | Very high |
| Renewal at standard terms | Repeated value and price acceptance | Existing-customer context | Strongest |
Move upward through the hierarchy. Do not use a survey result to claim validation when no qualified buyer will accept a proposal.
The decision comes before the method
Pricing research can answer different questions:
- Which customer segment has the strongest economic problem?
- Which alternatives create the relevant price anchor?
- Should the metric be seats, workspaces or usage?
- Which capabilities belong in each package?
- Is €49, €99 or €199 a credible starting range?
- Will buyers accept annual commitment?
- How much discount is required for prepayment?
- Why do current customers reject an increase?
- Which accounts would migrate badly to a new model?
Write the decision and uncertainty first.
A useful brief says:
We need to choose an initial package and price for agencies with 10–30 employees. We are uncertain whether buyers budget the product as team software or per-client delivery cost. Research must identify the budget owner, relevant alternatives, expected account size and behavioral response to a workspace package at €99–€249 per month.
A vague brief such as “find the optimal price” invites false precision.
Segmentation before aggregation
Willingness to pay varies because customers differ. Useful segmentation dimensions include:
company size, how often the use case recurs, how severe the problem is, what alternative is already in place, whether the context is regulated, whether the purchase is self-serve or sales-assisted, whether use is professional or occasional, usage volume, geography and currency, how mature the customer's process is, and how much money is at risk when it fails.
Not all of these separate buyers equally. Severity of the problem and the financial value at risk usually split a market more sharply than company size, which is the dimension teams reach for first because it is the easiest to look up.
Do not mix students, freelancers, agencies and enterprise departments into one average if they buy for different outcomes.
Write recruitment criteria
For every study, define who qualifies.
Example:
- owns or influences the relevant budget;
- works at an agency with 10–30 employees;
- manages at least eight active client projects;
- experienced the target problem in the last month;
- evaluated or purchased a related tool in the last two years;
- can describe the current workflow and cost.
Exclude participants who only match a broad demographic label but lack purchase context.
Separate user and buyer research
Users explain workflows and value moments. Buyers explain budgets, alternatives and approval. In small businesses, one person may cover both. In larger organizations, interview both and compare their assumptions.
Start with problem and alternative interviews
Do not open with your price. First understand the decision environment.
Problem questions
- Tell me about the last time this problem occurred.
- What triggered action?
- Who was involved?
- What did the problem delay, cost or put at risk?
- How often does it happen?
- What happens if you do nothing?
- Which part is most expensive or frustrating?
- How do you know the problem is resolved?
Ask for a recent event rather than general opinion.
Alternative questions
- How do you solve this today?
- Which software, people or manual work are involved?
- What did you evaluate before choosing it?
- What does the current solution cost in fees and time?
- Which contract or commitment applies?
- What would make you replace it?
- What is good enough about doing nothing?
The real alternative is often a spreadsheet, employee, agency, delayed project or accepted risk—not a direct software competitor.
Budget and purchase questions
- Which budget pays for the current approach?
- Who owns that budget?
- How was the last comparable purchase approved?
- At what amount does another approver become involved?
- Is spend planned annually or available during the year?
- Does the buyer prefer operating expense, project spend or committed contract?
- Which evidence is needed for renewal?
These questions reveal commercial constraints without asking the respondent to invent a price.
The current economic situation in numbers
Build a transparent value model with the buyer.
Possible inputs include:
- hours spent per occurrence;
- number of occurrences;
- fully loaded labor or supplier cost;
- revenue delayed or lost;
- conversion difference;
- avoidable payment or infrastructure fees;
- expected loss and probability of risk;
- capacity created;
- cost of errors, rework or compliance incidents.
Use ranges:
| Input | Conservative | Expected | High |
|---|---|---|---|
| Incidents per month | 10 | 20 | 35 |
| Hours per incident | 0.5 | 1 | 1.5 |
| Loaded hourly cost | €30 | €45 | €60 |
| Monthly labor exposure | €150 | €900 | €3,150 |
Then test every assumption:
- Does saved time reduce spend or only create capacity?
- Is the outcome attributable to the product?
- How quickly is value realized?
- What implementation work reduces the first-year benefit?
- Who trusts the calculation?
A value model supports pricing; it does not entitle the vendor to capture all created value.
Present concrete offers, not abstract prices
A price has meaning only with a package.
Before asking anyone about price, specify what you are pricing: the intended customer, the promised outcome, included capabilities, the value metric, the allowance or limit, support and onboarding, contract period, billing timing, cancellation or refund terms, and the expected time to value.
An underspecified offer is what makes price research unreliable. People answer about the product they imagined, and you find out what they imagined only when you try to deliver it.
Present two or three credible alternatives when researching packaging. Do not create one obviously bad decoy solely to manipulate the answer.
Ask diagnostic follow-ups
After presenting an offer, ask:
- What is your first reaction?
- Which part requires explanation?
- How would you compare this with the current alternative?
- Which budget would pay?
- Who else needs to agree?
- What risk makes this difficult to approve?
- Which included item is unnecessary?
- What would you expect at a higher or lower package?
- What would you do next if this were available today?
Do not defend the price immediately. Silence often produces the most useful explanation.
Direct price questions: use carefully
A single “How much would you pay?” question is weak because:
- respondents try to negotiate;
- some try to please the researcher;
- the offer is underspecified;
- the relevant budget is unknown;
- people are poor at forecasting behavior;
- the first number anchors the discussion;
- there is no consequence for being wrong.
Better direct questions include:
- What do you spend on the current alternative?
- Which internal threshold changes the approval process?
- At what total cost would this no longer be a priority?
- Which amount could you approve without another stakeholder?
- What result would justify this amount at renewal?
Treat answers as hypotheses to test through behavior.
Van Westendorp Price Sensitivity Meter
The Van Westendorp method asks four questions about a defined offer:
- At what price would it be so cheap that you would question its quality?
- At what price would it feel like a bargain?
- At what price would it start to feel expensive but still worth considering?
- At what price would it be too expensive to consider?
Aggregated response curves are used to describe perceived price boundaries and intersections.
When it can help
- respondents understand the category;
- the package is concrete;
- buyers have relevant purchase experience;
- the sample is large and representative enough for descriptive analysis;
- you need a range for further testing rather than a final answer.
Limitations
- answers remain hypothetical;
- respondents may not know unfamiliar-category prices;
- wording and currency anchor results;
- “cheap” can represent quality concern or simply preference;
- mixed segments produce misleading intersections;
- the method does not model actual conversion or retention;
- a chart intersection is not an economically optimal price.
Report the sample, segment and question wording. Do not publish an exact “optimal price” unsupported by behavior.
Gabor–Granger price testing
Gabor–Granger presents a defined offer at a price and asks purchase likelihood. Depending on the response, the next amount is higher or lower. Aggregated intent can produce a stated demand and revenue curve.
When it can help
- candidate amounts are already plausible;
- the offer is understood;
- respondents are qualified buyers;
- the sample supports segmentation;
- you want to compare a limited price range.
Limitations
- stated intent overestimates behavior;
- price sequence can anchor responses;
- respondents may infer desired answers;
- purchase likelihood scales are interpreted inconsistently;
- modeled revenue ignores retention, margin and acquisition;
- testing price without package context is meaningless.
Randomize starting points where appropriate and validate the selected range in an actual offer test.
Conjoint and discrete-choice research
Conjoint-style research asks respondents to choose between offers with varying attributes and prices. It can estimate relative trade-offs among: package features, support levels, limits, contract terms, brand or service conditions and price.
It can be useful when packaging contains several important dimensions and the organization has enough sample, research expertise and decision value to justify the study.
It is not a shortcut for early-stage uncertainty. Poor attributes, unrealistic combinations or an unrepresentative panel produce precise-looking but unhelpful output. A founder with 15 qualified prospects often learns more from concrete proposal tests than from an improvised conjoint survey.
Behavioral pricing tests
Behavioral tests provide stronger evidence because the participant faces a real trade-off.
Paid concierge pilot
Deliver the outcome manually or with limited software. Charge for a defined scope.
Learn:
- whether the buyer can approve payment;
- which result matters;
- actual delivery and support cost;
- whether the package is complete;
- what creates renewal intent.
A discounted pilot should state the future standard terms. Otherwise it validates only the discounted pilot.
Proposal test
Send a written offer with scope, amount, timing and next step. Track: acceptance, negotiation, stakeholder involvement, time to decision, reason for loss and requested conditions.
Do not count “send me something” as acceptance.
Deposit or reservation
For a pre-launch product, a refundable deposit or paid design-partner agreement can test commitment. Explain delivery status, refund conditions and timeline truthfully.
Landing-page or checkout test
A real traffic experiment can compare conversion at different offers, but it requires:
- comparable traffic cohorts;
- enough observations;
- consistent package and messaging;
- ethical handling when delivery is not immediate;
- analysis of downstream activation and refund behavior.
A fake checkout that implies an available product and then refuses purchase damages trust. Use a transparent waitlist, reservation or pilot application.
Sales cohort test
Offer different prices to comparable new prospects during defined periods or randomized assignments. Control discount authority and record segment differences.
In low-volume B2B sales, qualitative context matters. One enterprise win does not prove a price, and one loss does not disprove it.
Existing-customer research
Existing customers provide rich usage and outcome evidence, but they are anchored to current terms.
Existing customers are the best available evidence. Study activation and the workflows that stuck, the distribution of usage, the outcome actually achieved, support burden, where package limits bind, expansion history, discounts and exceptions, renewal conversations, and gross margin by account.
Discounts and exceptions are the most honest series in that list. They record what your own sales team believed the price should be, one deal at a time.
Ask about value before discussing an increase
A useful sequence:
- What prompted the original purchase?
- Which workflows rely on the product now?
- What changed after adoption?
- Which capability would be hardest to replace?
- What alternatives would be considered today?
- Who evaluates renewal?
- Which evidence supports that decision?
- How would a proposed package or price change affect planning?
Do not threaten access to force a favorable interview answer.
Distinguish fairness from affordability
A customer can believe a new price is fair but lack current budget. Another can afford it but believe the change violates the original exchange.
Record separately: perceived value, affordability, predictability, transition fairness, contract constraints and switching intention.
Objections: analysis instead of a loose count
Create a coded objection log.
| Objection | Possible cause | Evidence to collect | Possible response |
|---|---|---|---|
| “Too expensive” | Weak value, wrong segment or high amount | Alternative, budget, expected outcome | Reposition, repackage or retest amount |
| “Need a monthly option” | Commitment risk | Cash flow, trust, purchase stage | Monthly premium or pilot |
| “Usage is unpredictable” | Weak metric controls | Historical range, peak behavior | Allowance, cap, alerts or bands |
| “Missing feature” | Package or product gap | Frequency by segment, commitment | Add, reject or create another package |
| “I need approval” | Wrong interview role | Buying process and threshold | Reach budget owner |
| “Not now” | Weak urgency or timing | Trigger, planning cycle, alternative | Change segment or follow-up event |
| “Send information” | Polite dismissal or process | Specific next step and stakeholder | Request scheduled review |
Code the first objection and the root cause when known. “Price” is often the easiest socially acceptable answer.
From research to a price range
Use four boundaries.
Customer-value ceiling
Estimate the value created and the portion a buyer can credibly share. Use conservative, expected and high scenarios.
Alternative anchor
Write down what the customer would otherwise do: pay a competitor, employ someone or hire a contractor, absorb a transaction fee, run a manual process, accept a delay, retain the risk, or do nothing.
Do-nothing is the alternative that competitive analyses leave out, and it is the one you are usually losing to when a deal simply stops moving.
Economic floor
Calculate the minimum sustainable price including:
- variable delivery;
- support and onboarding;
- payment fees;
- expected refunds and bad debt;
- acquisition and sales effort;
- required contribution to fixed costs;
- margin buffer for uncertainty.
Behavioral evidence
Use accepted proposals, pilots and renewals to narrow the range.
A viable price must sit above the economic floor, below a credible customer-value ceiling and inside a range buyers actually accept under realistic terms.
Do not optimize revenue per visitor alone
A higher price improves the modelled revenue immediately and can quietly reduce activation quality, retention, expansion, referrals, sales velocity, cash collection — and gross margin itself, once higher expectations bring more support with them.
That last one surprises people. Customers paying more ask for more, and a price rise that arrives without a service change funds itself out of your own team's time.
A lower price can increase conversion while attracting low-fit customers and making acquisition uneconomic.
Evaluate the full system:
Qualified demand × paid conversion × retained revenue × gross margin − acquisition and service cost
Use cohort data when available.
A two-week research sprint
Day 1: write the decision brief
Define: target segment, pricing decision, current hypotheses, known constraints, evidence threshold, owner and decision date.
Days 2–3: recruit qualified participants
Recruit users, buyers and budget owners with recent problem or purchase experience. Avoid relying only on friends, existing fans or low-cost survey panels.
Days 4–7: run eight to twelve interviews
Cover recent problem, alternatives, value, budget and buying process. Present concrete offers only after understanding context.
Capture notes in a consistent template and mark direct quotes separately from interpretation.
Day 8: synthesize segments and objections
Group participants by behavior and buying context. Identify repeated evidence and unresolved disagreement.
Days 9–10: build price and package options
Create two or three offers supported by value, alternative and cost evidence. Model unit economics and edge cases.
Days 11–13: request commitment
Present written pilot or standard offers to qualified prospects. Ask for an explicit next action with a date.
Day 14: decide the next test
Document the decision: the selected segment, the package, the value metric, the price or range, billing terms, the evidence that supports it, the evidence against it, the assumptions still open, and the thresholds for success and stopping.
The contrary-evidence field is what makes the document worth keeping. A pricing decision recorded with only supporting evidence cannot be revisited, only defended.
The sprint should end with an experiment, not a permanent proclamation.
Research quality and ethics
Do not deceive participants
State whether the product is live, in development or manually delivered. Do not imply scarcity, customer logos or outcomes that do not exist.
Handle incentives carefully
Research compensation pays for time, not favorable answers. Separate the incentive from any purchase negotiation.
Protect confidential data
Do not ask participants to disclose sensitive employer budgets or contracts they are not permitted to share. Store recordings and notes with appropriate consent and access control.
Report uncertainty
Distinguish: observed behavior, participant statement, researcher interpretation and modeled assumption.
Do not turn eight interviews into a claim about an entire market.
Avoid discriminatory or exploitative pricing
Segment pricing by legitimate differences such as package, cost, volume, geography or service—not protected characteristics or opaque manipulation. Make material terms understandable.
Metrics after launch
Purchase
- Qualified offer-to-paid conversion
- Win rate by segment and amount
- Sales cycle
- Discount and exception rate
- Checkout abandonment
- Reason for loss
Value
- Activation
- Time to first value
- Retained usage
- Outcome attainment
- Support burden
Revenue quality
- Gross and net revenue retention
- Expansion and contraction
- Refund and dispute rate
- Payment failure
- Revenue concentration
Economics
- Gross margin by package and segment
- Contribution margin
- CAC payback
- Service hours per account
- Cash collection timing
Review price together with package, metric and customer quality.
Common research mistakes
Asking only current users
Current users survived acquisition and onboarding. Include lost prospects, qualified non-buyers and relevant new buyers.
Mixing incompatible segments
An average across different buying systems is not a useful target price. Analyze segments separately before deciding whether one offer can serve them.
Showing the price too early
Early anchoring changes the problem and value conversation. Understand the context first.
Treating stated intent as conversion
“Probably would buy” is not revenue. Ask for the next commitment.
Using a survey without a concrete package
Respondents cannot evaluate a number without scope, limits, terms and outcome.
Ignoring delivery cost
High willingness to pay does not guarantee a viable price when onboarding, compute or service work is expensive.
Selecting the method because it creates a chart
A polished curve cannot rescue poor recruitment or a hypothetical offer. Choose the cheapest method capable of reducing the decision's uncertainty.
Continuously interviewing without deciding
Set a decision date and evidence threshold. Research should support action and further testing.
Decision checklist
Research design
- The pricing decision and target segment are explicit.
- Participants have relevant problem and buying experience.
- User, buyer and payer roles are represented where distinct.
- The offer includes package, metric, amount and terms.
- Sample limitations are documented.
Evidence
- Current alternatives and spend are understood.
- Value assumptions use conservative ranges.
- Buyers reacted to concrete offers.
- At least some evidence involves real commitment.
- Contradictory evidence is retained rather than discarded.
Economics
- Variable and human delivery costs are included.
- Margin is modeled by segment and usage.
- Acquisition and sales effort fit the price.
- Payment timing and discounts are modeled.
Decision
- The selected price has a documented rationale.
- Package and metric are not confused with amount.
- Success, revision and stop thresholds are defined.
- Post-launch metrics can detect poor-fit customers or margin.
- The next review date is scheduled.
Ask about behaviour, not opinions
Treat willingness to pay as behavior in a specific buying context, not an answer to a hypothetical question.
Interview qualified users and buyers to understand problems, alternatives, budgets and risk. Use survey methods to structure uncertainty when the sample supports them. Then present a concrete offer and ask for a real commitment.
The best initial price is not the number that produces the most enthusiastic research response. It is a defensible hypothesis that qualified customers will pay, the product can deliver profitably and the team can revise as stronger evidence arrives.