Customer-Focused Competitive Pricing Research: A Founder's Playbook
Direct answer: Use Van Westendorp to establish an acceptable price corridor, then use Gabor-Granger to test demand and modeled revenue within that corridor. Treat the result as stated purchase intent and simulated revenue until an in-market test provides observed behavior. Add MaxDiff or conjoint only when feature or package trade-offs materially affect the decision.
The central decision is not which method sounds most sophisticated. It is which method answers the pricing question in front of the team. Competitor monitoring can show what other companies publish, but customer-focused research asks how the intended buyer perceives value and responds to price. Those are different evidence sources and should not be treated as interchangeable.
Match the pricing decision to the research method
Start with the uncertainty you need to resolve.
- If the acceptable price range is unknown, use Van Westendorp. The method identifies four price-perception intersection points and an acceptable price corridor, but it does not directly produce a demand curve or revenue estimate (Van Westendorp produces these four boundary points).
- If you have a plausible range and need to compare specified prices, use Gabor-Granger. It tests price points, builds a demand curve from purchase intent, and estimates modeled revenue by combining price with purchase probability (Gabor-Granger pricing method).
- If the decision is primarily about which features matter, use MaxDiff as a feature-value ranking exercise.
- If the decision involves packages, features, and price trade-offs together, consider conjoint. Conjoint can estimate feature trade-offs and simulate outcomes across package and price configurations, but it requires more complex design and analysis than simpler willingness-to-pay methods (Conjoint and willingness-to-pay comparison).
Figure 1. Decision map: Match the pricing question to the research method Alt text: Decision map connecting four SaaS pricing decisions to Van Westendorp, Gabor-Granger, MaxDiff, and conjoint.
Unknown acceptable price range → Van Westendorp Known range and price optimization → Gabor-Granger Feature prioritization without price → MaxDiff Package and price trade-offs → conjoint
The method should follow the decision. A founder with an uncertain price range usually benefits from a sequential Van Westendorp plus Gabor-Granger study, rather than beginning with a more elaborate package design.
Understand what each method can and cannot tell you
Van Westendorp produces a corridor for price perception. It is useful for framing the range to test, but its output is not a demand curve and not a revenue estimate. Gabor-Granger is more direct about specified prices: respondents provide purchase-intent responses at tested points, which can be used to construct a demand curve and a modeled revenue view.
The distinction between evidence layers matters. A response to a survey is stated purchase intent. A calculation that combines a tested price with a purchase probability is modeled or simulated revenue. A completed purchase, conversion, or renewal in the market is observed behavior. These labels should remain separate in a board deck and in the analysis. Modeled revenue is simulated until verified in market (stated intent, modeled revenue, and observed conversion).
Gabor-Granger presentation also requires a design choice. Sequential presentation can make a survey easier to follow, while randomized presentation can address a different set of concerns. Sequential and randomized designs each carry trade-offs, including anchoring risk and survey length (Sequential and randomized designs each carry trade-offs). The choice belongs in the method record, not hidden inside the results.
Figure 2. Process map: From buyer segment to pricing evidence Alt text: Process map showing a customer-focused pricing study from segment definition through market verification.
Define decision and segment → screen and recruit matched respondents → collect and clean responses → analyze method-specific outputs → separate intent, simulation, and behavior → plan market verification.
Conjoint is appropriate when the decision depends on several attributes at once. Its ability to simulate package and price configurations comes with more complex design and analysis than simpler willingness-to-pay methods. That complexity is justified only when those trade-offs are central to the decision.
Design and recruit a study you can trust
Define the segment before drafting questions. Write down the buyer context, the product being evaluated, and the decision the research must support. Screening should remove respondents who do not match the intended customer context. Recruitment should seek matched participants rather than treating a broad audience as a substitute for the target segment.
Keep the instrument focused. Ask only what is needed to answer the pricing decision, and make every price point explicit about its unit and billing period. For example, write “monthly price in USD” or “annual price in USD,” rather than presenting an unlabeled number. Use concise, neutral wording and document the order in which prices appear.
For Gabor-Granger, select price points that represent the range under consideration, then decide whether presentation will be sequential or randomized. Record that choice and its rationale because the design can affect anchoring risk and survey length. The pricing survey question guidance can help structure the wording, but the questions still need to match the defined segment and decision.
Clean responses using rules established before reviewing the outcome. Remove responses that fail the screen or do not provide usable answers, and keep a record of exclusions. Do not describe fielding speed, sample size, or timing as universal guarantees. They are study-design choices and depend on the segment, recruitment approach, and method.
Convert survey responses into a decision, not a forecast
For Gabor-Granger, the modeled revenue logic is straightforward: for each tested price, combine the price with the corresponding purchase probability to create a modeled revenue value. That calculation is useful for comparing scenarios, but it remains simulated. The purchase probability comes from stated intent, not observed conversion.
Illustrative example. Simulated: a team could compare several explicitly labeled monthly price points, pair each with its stated purchase probability, and rank the resulting modeled revenue scenarios. The example does not establish an actual conversion rate, realized revenue, or forecast. Any calibration assumption should be displayed beside the calculation rather than presented as a measured fact. The pricing analysis guidance can support the structure of this comparison.
Use the output to make a decision such as selecting a price corridor for verification, identifying a small set of prices for market testing, or deciding that the evidence is too uncertain to move forward. Then define the in-market check that will compare survey expectations with observed behavior.
Figure 3. Curve annotated range: Separate stated intent from modeled revenue Alt text: Annotated pricing curve showing tested price points, purchase intent, simulated modeled revenue, and a later observed-conversion check.
Illustrative example. Simulated: tested price points → purchase-intent response → simulated modeled revenue → calibration assumption → observed-conversion test.
Choose between DIY, self-serve software, and consulting
The execution path should reflect the decision's complexity and the team's research experience. DIY work offers control over wording, recruitment, and analysis, but requires the team to manage each step. Self-serve software can provide a structured workflow for a defined method. External consulting expertise may be appropriate when the design involves complex trade-offs, unfamiliar analysis, or a need for additional research support.
Compare the options by customization, research experience, speed, complexity, and cost considerations. Do not treat a tool choice as evidence quality by itself. The quality of the segment definition, questions, price design, cleaning rules, interpretation, and verification plan still matters.
Kinetic Pricing lists one-time methods including Price range finder at $149.00 USD per study, Price point tester at $199.00 USD per study, Feature value ranker at $279.00 USD per study, and Package and price builder at $499.00 USD per study. Kinetic Pro is listed at $99.00 USD per month or $990.00 USD per year, with a 30-day trial. These are product references, not evidence that a method will produce a particular business outcome. See Kineticpricing's plans and Kinetic Pro for the available paths.
Account for bias before presenting results
Interpretation: Present the study as decision support, not as a guaranteed forecast. The interpretation should state which responses represent stated intent, which calculations represent simulated modeled revenue, and which results, if any, represent observed conversion.
Limits: Before rollout, review four limitations. Hypothetical bias can separate what someone says from what they do. Anchoring can affect responses to presented prices, making the sequential or randomized design choice important. Range misspecification can leave the tested prices poorly aligned with the real decision. Sample nonrepresentativeness can make results less relevant to the intended segment. These are reasons to document assumptions and verify the selected price in market, not reasons to relabel survey data as behavior.
Figure 4. Limitation sequence: Limitations to address before rollout Alt text: Limitation sequence showing four pricing-research biases followed by an in-market verification step.
Hypothetical bias → anchoring → range misspecification → sample nonrepresentativeness → in-market verification.
The final evidence layer is observed conversion. Until that layer exists, state the boundary clearly: the research reports what respondents said, the model reports a simulated scenario, and the market test reports what customers did.
Turn the study into a stakeholder-ready next step
Method: Define the pricing decision and buyer segment, choose Van Westendorp, Gabor-Granger, MaxDiff, or conjoint according to that decision, recruit relevant respondents, collect and clean responses, and analyze the method-specific output.
Next step: Use a four-step sequence. First, define the decision and segment. Second, draft the questions and explicit price points. Third, recruit and clean matched responses. Fourth, model the result, label simulations, and verify the selected price with observed market behavior.
The recommended decision for an uncertain SaaS price is sequential: begin with Van Westendorp for the acceptable corridor, follow with Gabor-Granger for specified prices and modeled revenue, and add MaxDiff or conjoint only when feature or package trade-offs materially affect the choice. Kineticpricing's pricing research resources provide a starting point for the research path.
Additional context on these methods is available from external consulting expertise, Van Westendorp vs Gabor-Granger: Which to Choose | Quali-Fi and Gabor-Granger Pricing: Method, Steps & Van Westendorp Comparison | TestFeed.
Run this method with your users
Start 30-day free trial to run every method with your users, or Buy one study when one decision needs evidence now.
Sources
Quali-Fi, “Van Westendorp vs Gabor-Granger: Which to Choose”
TestFeed, “Gabor-Granger Pricing: Method, Steps & Van Westendorp Comparison”