Value-Based Pricing and Willingness-to-Pay Research

Introduction

Pricing is one of the most consequential decisions in brand management. It determines revenue, margin, market access, customer expectations and the perceived position of a brand. Yet pricing is often treated too narrowly—as a cost-plus calculation or as a response to competitors. A stronger approach begins with customer value: what benefits customers perceive, how strongly they need them, what alternatives cost, and how much they are prepared to sacrifice to obtain the offering.

Value-based pricing sets price in relation to the value delivered to a defined customer segment rather than simply adding a margin to cost. Willingness to pay (WTP) is the maximum amount a customer is prepared to exchange for an offering under specified conditions. WTP is not identical to satisfaction, affordability or observed purchase price. It is a behavioural and economic construct that varies with context, information, usage occasion, alternatives, brand trust and perceived risk.

This chapter explains how managers can estimate economic value and WTP using conjoint analysis, Van Westendorp price-sensitivity research, Gabor–Granger questions, controlled experiments and customer evidence. It also shows how to translate imperfect research findings into defensible price points.

Learning Objectives

By the end of this chapter, you should be able to:

1. Distinguish cost-based, competition-based and value-based pricing.
2. Explain economic value, perceived value and willingness to pay.
3. Select an appropriate WTP research method for a managerial problem.
4. Interpret conjoint, Van Westendorp and Gabor–Granger results critically.
5. Combine survey findings, behavioural evidence and financial constraints.
6. Recommend a price point, range or architecture that is commercially defensible.
7. Identify validity, ethical and governance risks in pricing research.

1. The Logic of Value-Based Pricing

A product creates value when its benefits exceed the customer’s total sacrifice. Benefits may include functional performance, convenience, emotional reward, social meaning, risk reduction and future flexibility. Sacrifice includes monetary price, time, effort, learning costs, switching costs and perceived risk.

Three broad pricing orientations are useful:

Cost and competition remain constraints, but they should not automatically determine price. A firm may have low costs but weak customer value, or high costs but a differentiated proposition. Value-based pricing asks what the offering is worth to a particular segment and how much of that value the firm can capture while maintaining adoption and trust.

Economic value and value communication

Economic value to the customer (EVC) compares an offering with the next-best alternative. A simplified expression is:

Here, P_ref is the price of the reference alternative, V_diff is the monetary value of additional benefits, and C_diff is the monetary equivalent of additional disadvantages or switching costs. All terms must use the same currency and time basis. The formula is most useful when the reference alternative is identifiable and benefit differences can be estimated credibly.

EVC is not automatically the final price. The firm may capture only part of the value because of competition, uncertainty, bargaining power, adoption barriers or strategic objectives. A practical price must also cover relevant incremental cost and support the desired brand position.

Willingness to pay

For a customer, WTP can be represented as the highest price at which the perceived utility of buying is at least as great as the utility of the next-best option, including not buying. In practice, WTP is conditional: “How much would this customer pay for this product, with these features, at this time, through this channel, compared with these alternatives?”

WTP distributions are more useful than a single average. A mean can conceal distinct segments: price-sensitive users, mainstream users and high-value users. Managers should therefore examine medians, percentiles, response distributions and segment differences.

2. From Value Drivers to Price Metrics

Before asking about price, define the value proposition. A value-driver map links product attributes to customer outcomes:

Attribute → customer benefit → economic or psychological outcome → evidence of value

For example, a cloud accounting platform may provide automated reconciliation. The benefit is fewer manual errors; the outcome may be reduced labour time and lower compliance risk; evidence may include time records, error rates and renewal behaviour.

A useful distinction is:

• Feature: what the product has or does.
• Benefit: what the customer gains.
• Outcome: the measurable result for the customer.
• Value metric: the unit by which customers perceive or are charged for value, such as users, transactions, usage volume or outcomes achieved.

Pricing by a value metric is attractive when the metric is understandable, measurable, scalable and closely related to customer value. Poor value metrics create perceived unfairness, unpredictable bills or incentives to avoid usage.

Price thresholds and price architecture

A price point is a specific amount, whereas a price range reflects uncertainty. A price architecture includes list price, discounts, bundles, tiers, payment terms, usage charges and upgrade paths. Research should therefore test not only “the price” but also the commercial structure.

Important metrics include:

These formulas assume consistent units and that price and cost are measured on the same transaction basis. If quantity changes with price, the manager must estimate demand rather than treating quantity as fixed. A high-WTP price may reduce volume; a lower price may increase volume but reduce contribution per unit.

3. Research Design: Asking the Right Question

WTP research fails when the research question is vague. A sound design specifies:

1. The decision: launch price, repositioning, tier design, discount policy or packaging.
2. The target population and buying unit.
3. The relevant competitive alternatives.
4. The purchase context and time horizon.
5. The benefit description and level of product information.
6. The decision metric: revenue, contribution, penetration, retention or brand positioning.

The sample should represent actual or likely buyers, not merely convenient respondents. Screening should establish category relevance, decision authority, usage frequency and purchase timing. Researchers should distinguish users from economic buyers in business markets.

A questionnaire should avoid leading claims such as “excellent value” and should make payment conditions explicit. Prices should state currency, billing period, taxes, delivery fees and contract terms. For a subscription, monthly and annual prices should not be mixed without explanation.

4. Conjoint Analysis

Conjoint analysis estimates how respondents trade off product attributes and price. Instead of asking directly what each feature is worth, it presents realistic alternatives and infers preference patterns from choices or ratings.

Structure

A conjoint study defines:

• Attributes, such as battery life, delivery speed or warranty.
• Levels, such as 24, 48 or 72 hours.
• Alternatives, constructed from combinations of levels.
• Choice tasks, in which respondents select an option, including “none” where appropriate.

A simple random utility representation is:

where U_ij is the utility respondent i receives from option j, V_ij is systematic utility explained by observed attributes, and ε_ij is unobserved influence. Respondents are assumed to choose the option with the highest utility. The model is probabilistic because unobserved influences vary.

For a linear specification:

The β coefficients represent utility weights, and x values code attribute levels. Coding choices matter: effects coding and dummy coding produce different coefficient interpretations, although predicted utilities can be equivalent when specified correctly.

If price has coefficient β_p and a feature has part-worth β_f, the feature’s implied monetary value is:

This calculation is valid when utility is linear in price, β_p < 0, both coefficients use compatible coding, and the ratio is interpreted within the tested range. If price effects are nonlinear or the coefficient varies by person, the ratio is an approximation rather than a universal value.

Worked example

Suppose a choice model estimates a part-worth of 1.20 utility units for express delivery and a price coefficient of −0.04 utility units per currency unit. Then:

The estimated average incremental WTP is 30 currency units, subject to the model’s assumptions. It does not mean every customer will pay 30, nor does it justify adding 30 to any arbitrary price.

Strengths and weaknesses

Conjoint analysis is strong for feature trade-offs, product configuration, segmentation and competitive simulation. It can estimate the effect of price relative to attributes and can include a no-purchase option.

Limitations include hypothetical bias, respondent fatigue, unrealistic attribute combinations, poor comprehension and model dependence. A choice task is not a real transaction. If the tested price range is too narrow, the model may extrapolate dangerously. Researchers should validate predictions with market behaviour whenever possible.

5. Van Westendorp Price Sensitivity Meter

The Van Westendorp method asks four threshold questions:

1. At what price is the product so cheap that quality would be questioned?
2. At what price is it a bargain?
3. At what price is it becoming expensive, but still worth considering?
4. At what price is it too expensive to consider?

Responses are converted into cumulative distributions. The intersections of selected curves produce indicators such as the point of marginal cheapness, point of marginal expensiveness, acceptable price range and often a “point of optimum price” under the method’s terminology.

The method is useful for early exploration of perceived price boundaries, particularly when a product is new and competitive alternatives are unclear. It is relatively simple and provides a visual conversation about price perceptions.

However, it does not present realistic purchase choices and does not directly estimate demand, revenue or contribution. Respondents may give internally inconsistent answers, and stated thresholds can be influenced by framing, reference prices and brand knowledge. The acceptable range should therefore be treated as a diagnostic interval, not a mechanically correct price.

6. Gabor–Granger Price Testing

Gabor–Granger research presents a product description and asks purchase intention at one or more prices. The sequence may be randomized or designed to reduce anchoring. The resulting proportion indicating purchase at each price creates an estimated price-response curve.

If q(p) is the proportion of respondents willing to buy at price p, a simple expected-revenue index is:

This is an index when the sample proportion is used; it becomes expected revenue per potential customer only under the assumption that the sample represents the target population and the purchase horizon is defined. It does not include variable cost, repeat purchases, acquisition cost or channel margin.

For contribution per potential customer, a simplified expression is:

where c is variable cost per unit. If costs, repeat rates or channel commissions vary by price, they must be incorporated separately.

Worked example

A survey produces the following purchase proportions:

The highest revenue index is 17.00 at a price of 25. If variable cost is 10, contribution indices are 8.00, 10.20, 10.40 and 8.75 respectively. Under these assumptions, 30 maximizes contribution among the tested prices, even though 25 maximizes revenue.

Gabor–Granger is accessible and useful for a small number of price points, but purchase intention is not actual purchase. Sequential exposure can create anchoring, and respondents may overstate willingness because no money is at risk. Randomized price presentation, realistic competitive context and behavioural validation improve credibility.

7. Experiments and Customer Evidence

The strongest evidence usually comes from behaviour. Possible approaches include:

• A/B tests of prices or packages.
• Geo-experiments with treatment and control markets.
• Landing-page tests using genuine purchase or deposit behaviour.
• Pilot launches with pre-specified guardrails.
• Sales-force quote data and win–loss analysis.
• Renewal, downgrade and churn data.
• Customer interviews about alternatives, budgets and consequences.
• Transaction data showing price, quantity, segment and channel.

A basic price experiment compares outcomes under different prices while controlling, as far as possible, for audience, timing, promotion, channel and inventory. Random assignment helps make differences attributable to price rather than selection. The analyst should monitor conversion, average order value, contribution, refunds, complaints, retention and brand effects—not conversion alone.

A useful causal distinction is:

• Observed correlation: high-value customers may pay more because they are already different.
• Causal price effect: changing price causes a measurable change in behaviour under comparable conditions.

Ethically, experiments must not create misleading claims, discriminatory access or unexpected charges. Customers should receive clear terms, and essential services require particular care.

8. Integrating Research into a Defensible Price Point

No single method reveals the “true” price. Each method measures a different object:

A defensible pricing process can follow six steps:

1. Define value and alternatives. Identify the reference option, customer outcomes and value metric.
2. Establish financial boundaries. Calculate variable cost, required contribution, channel economics and capacity constraints.
3. Explore perceptions. Use interviews and Van Westendorp to identify language, thresholds and possible ranges.
4. Model trade-offs. Use conjoint when features, bundles or segments matter; use Gabor–Granger for a focused price-response estimate.
5. Validate behaviour. Conduct an experiment, pilot or analysis of real transactions.
6. Select and govern the decision. Choose a price or architecture, document evidence and define monitoring triggers.

The final recommendation should show a range and a rationale, for example: “Recommend 30 currency units because it lies within the perceived acceptable range, captures the targeted feature value, maximizes estimated contribution among tested prices, and remains consistent with the premium positioning. Validate through a controlled market pilot.”

Managers should use sensitivity analysis. If the recommended price changes substantially under plausible assumptions about conversion, cost or retention, the decision is uncertain and should be treated as a testable hypothesis rather than a fact.

9. Segmentation, Fairness and Brand Meaning

WTP heterogeneity creates opportunities for differentiated packaging, but not every observed difference justifies different prices. Segment-specific pricing should reflect differences in service cost, value received, usage, willingness to commit or legitimate market conditions. It should not rely on protected characteristics or exploit vulnerable consumers.

Price communicates brand meaning. A very low price may undermine quality inference; a premium price can support exclusivity but raises performance expectations. Discounts can train customers to wait, weaken reference price integrity and reduce trust. A coherent architecture may use good–better–best tiers, transparent bundles, annual commitment discounts or usage-based pricing.

Reference price is the price customers expect or remember. It is shaped by competitors, past purchases, promotions and category conventions. Managers should monitor not only WTP but also price fairness: whether customers believe the price is justified by value and applied consistently.

10. Illustrative Case: TerraBrew Office Coffee System

Background

TerraBrew, a fictional small appliance brand, plans to launch an office coffee system for organizations with 20–100 employees. The system reduces preparation time, offers quality control and includes a service subscription. Management must choose equipment and subscription prices without damaging the brand’s sustainable-premium position.

Problem

The team initially considered a cost-plus price of 240 currency units for the machine and 18 per employee per month. Sales interviews suggested that office managers valued reduced administration, while finance managers focused on predictable total cost. Competitors charged between 190 and 320 for equipment, but their service levels differed.

Analysis and approach

Interviews identified three outcomes: fewer supply interruptions, lower administrative time and more consistent quality. A Van Westendorp study suggested that the machine’s perceived acceptable range was approximately 210–290, but respondents were uncertain because the category was unfamiliar. A conjoint study tested equipment price, service response time, sustainability certification and subscription level. The model indicated that rapid service and certification had meaningful value, but price sensitivity varied sharply by office size.

Gabor–Granger testing produced the following estimated purchase proportions for the machine among qualified buyers:

The revenue index slightly favoured 250. With variable fulfilment cost of 120, contribution indices were 62.00, 71.50, 73.60 and 64.60. Thus, 280 produced the highest estimated contribution among tested prices, but the difference from 250 was not decisive given survey uncertainty.

The firm then ran a two-city pilot with prices of 250 and 280, using comparable sales territories and identical service terms. It pre-specified monitoring of conversion, contribution, service complaints and 90-day activation. The higher price reduced conversion but increased contribution per sale; activation and complaint rates were similar. Management selected 280 for the premium service package and introduced a 250 package with slower standard service rather than simply discounting the same offer.

Outcome and learning points

The case demonstrates that price should be linked to the value proposition and package design. Survey methods helped identify thresholds and trade-offs, while the pilot tested real behaviour. The final decision was not “the price discovered by research”; it was a governed managerial choice supported by converging evidence.

Key lessons are:

• A competitor’s price is not necessarily the correct reference price.
• Revenue and contribution can favour different prices.
• Feature differentiation can make a price increase credible.
• Segment and package design may be more effective than blanket discounting.
• Pilot results should include customer and brand outcomes, not only immediate sales.

11. Limitations, Ethics and Governance

WTP studies are vulnerable to hypothetical bias, strategic responses, sample imbalance, framing effects, anchoring, order effects and nonresponse. Researchers should disclose uncertainty and avoid false precision. Confidence intervals or credible intervals, where appropriate, should accompany estimates; a point estimate without uncertainty can encourage overconfident decisions.

Data governance requires lawful collection, purpose limitation, access control, retention rules and responsible use of customer-level information. Sensitive attributes should not be used to create exploitative price differences. Dynamic pricing must be transparent enough for customers to understand material price differences, particularly in essential or high-stakes categories.

AI can assist with text coding, demand forecasting, segmentation and experimental analysis. However, automated recommendations may reproduce historical discrimination, confuse correlation with causation or optimize short-term revenue at the expense of trust. Human review, model documentation, fairness checks and monitoring are essential. AI-generated price recommendations should be treated as decision support, not authority.

12. Contemporary Developments and Best Practice

Digital commerce enables rapid price testing, personalized offers and subscription optimization. These capabilities increase the value of experimentation but also increase governance risk. Subscription businesses increasingly monitor acquisition price, retention, expansion, downgrade and lifetime contribution rather than one-time WTP.

Best practice includes:

• Define the customer and buying context before selecting a method.
• Use a real reference alternative and realistic product description.
• Test price as part of a complete offer, including terms and service.
• Include a no-purchase option in choice research where appropriate.
• Randomize question order and price exposure when feasible.
• Separate exploratory research from confirmatory testing.
• Combine stated preference with transaction evidence.
• Report segment variation and uncertainty.
• Optimize contribution and long-term value, not conversion alone.
• Document assumptions, exclusions and decision rights.
• Monitor fairness, complaints, retention and brand effects after launch.

Chapter Summary

Value-based pricing relates price to customer-perceived and economically relevant value rather than relying exclusively on cost or competition. Economic value to the customer compares an offering with a reference alternative, while WTP represents the maximum conditional sacrifice a customer will make.

Conjoint analysis estimates attribute and price trade-offs; Van Westendorp identifies perceived price thresholds; Gabor–Granger estimates stated purchase response across prices; and experiments provide stronger behavioural evidence. Each method has a distinct purpose and bias. The most defensible price decision triangulates methods, incorporates cost and channel economics, recognizes segment heterogeneity, and validates assumptions through real-market evidence.

Pricing is also a brand decision. It communicates quality, fairness and positioning. Ethical governance requires transparency, protection of customer data, scrutiny of algorithmic discrimination and avoidance of exploitative practices. A recommended price should therefore be presented as a reasoned, testable decision supported by evidence and explicit uncertainty.

Key Terms

Value-based pricing: Setting price in relation to customer-perceived value and outcomes.

Willingness to pay: The maximum price a customer is prepared to pay under specified conditions.

Economic value to the customer: The value of an offering relative to a reference alternative.

Reference alternative: The next-best option against which an offering is evaluated.

Perceived value: The customer’s assessment of benefits relative to total sacrifice.

Value metric: The unit used to measure or charge for delivered value.

Price architecture: The structure of prices, tiers, bundles, discounts and payment terms.

Conjoint analysis: A method that estimates preferences from trade-offs among product attributes and price.

Part-worth utility: The estimated utility contribution of an attribute level.

Random utility model: A model separating systematic utility from unobserved influences.

Van Westendorp Price Sensitivity Meter: A threshold-based method for exploring perceived acceptable prices.

Gabor–Granger method: A stated-intention method measuring purchase response at different prices.

Price-response curve: The relationship between price and purchase or demand probability.

Hypothetical bias: The tendency for stated intentions to differ from actual behaviour.

Reference price: The price a customer expects based on experience, comparison or memory.

Contribution: Revenue remaining after variable costs, before or after specified fixed costs depending on the accounting definition used.

Price experiment: A controlled test of different prices or offers using behavioural outcomes.

Price fairness: The perceived legitimacy and equity of a price.

Price discrimination: Charging different prices for materially similar offers under defined conditions.

Triangulation: Combining multiple forms of evidence to improve decision credibility.

Review Questions

Short Answer Questions

1. What distinguishes value-based pricing from cost-plus pricing?
2. Define willingness to pay and explain why it is conditional.
3. What are the four standard Van Westendorp questions designed to identify?
4. Why might revenue-maximizing and contribution-maximizing prices differ?
5. State one major strength and one limitation of conjoint analysis.

Long Answer Questions

1. Explain how a firm can move from a value-driver map to a defensible price architecture.
2. Compare conjoint analysis, Van Westendorp and Gabor–Granger in terms of purpose, data and limitations.
3. Discuss how experiments can improve on stated WTP research.
4. Explain why segmentation is important when interpreting WTP distributions.
5. Evaluate the ethical and governance issues associated with personalized and dynamic pricing.

Analytical / Application Questions

1. A product has a variable cost of 12. At prices of 20, 25 and 30, estimated purchase proportions are 0.70, 0.55 and 0.38. Calculate the contribution index at each price and recommend which tested price maximizes it.
2. A conjoint model gives a feature part-worth of 0.90 and a price coefficient of −0.03 utility units per currency unit. Estimate the feature’s implied WTP and state two assumptions behind the calculation.
3. Design a research programme for launching a premium subscription service. Explain which methods you would use at exploratory, modelling and validation stages, and identify at least three risks to control.

Multiple Choice Questions with Answers

Question 1

Which statement best describes value-based pricing?

A. It sets price by adding a fixed margin to accounting cost.

B. It sets price solely equal to the market leader’s price.

C. It relates price to the value delivered to a defined customer segment.

D. It uses the lowest possible price to maximize market share.

Correct Answer: C. It relates price to the value delivered to a defined customer segment.

Question 2

In the simplified economic value expression EVC = P_ref + V_diff − C_diff, what does P_ref represent?

A. The firm’s fixed cost

B. The price of the reference alternative

C. The customer’s total annual income

D. The price discount offered by the firm

Correct Answer: B. The price of the reference alternative

Question 3

A conjoint model estimates a feature part-worth of 0.80 and a price coefficient of −0.04. What is the feature’s implied WTP under the linear-price assumption?

A. 5 currency units

B. 20 currency units

C. −20 currency units

D. 32 currency units

Correct Answer: B. 20 currency units

Question 4

Which method primarily identifies perceived price thresholds such as bargain and too-expensive levels?

A. Van Westendorp Price Sensitivity Meter

B. Randomized controlled trial

C. Inventory turnover analysis

D. Break-even analysis

Correct Answer: A. Van Westendorp Price Sensitivity Meter

Question 5

If price is 25 and the estimated purchase proportion is 0.60, what is the Gabor–Granger revenue index?

A. 15.00

B. 25.60

C. 41.67

D. 0.024

Correct Answer: A. 15.00

Question 6

Why can a contribution-maximizing price be higher than a revenue-maximizing price?

A. Contribution ignores variable cost.

B. A higher price can produce more contribution per sale despite lower volume.

C. Revenue always excludes price.

D. Higher prices necessarily increase quantity sold.

Correct Answer: B. A higher price can produce more contribution per sale despite lower volume.

Question 7

What is the principal advantage of a real price experiment over a purely stated WTP survey?

A. It eliminates all sampling error.

B. It measures behaviour under an actual offer.

C. It requires no ethical review.

D. It guarantees the optimal long-term price.

Correct Answer: B. It measures behaviour under an actual offer.

Question 8

Which item is a value metric?

A. A customer’s favourite colour

B. The number of users charged under a software plan

C. The company’s office rent

D. The survey completion time

Correct Answer: B. The number of users charged under a software plan

Question 9

What is a major limitation of Van Westendorp research?

A. It cannot ask respondents about price perceptions.

B. It directly measures actual transaction demand.

C. It does not by itself estimate demand, revenue or contribution.

D. It requires a fully operational product.

Correct Answer: C. It does not by itself estimate demand, revenue or contribution.

Question 10

Which practice most improves the defensibility of a pricing recommendation?

A. Relying on one average WTP estimate

B. Ignoring variable costs to focus on adoption

C. Combining stated research with behavioural evidence and documenting assumptions

D. Choosing the highest observed competitor price

Correct Answer: C. Combining stated research with behavioural evidence and documenting assumptions