Conjoint analysis: product trade-offs with examples

Conjoint analysis estimates what people value by asking them to compare complete product offers. Varying features and prices across choices reveals the trade-offs behind their preferences.

What is conjoint analysis?

Conjoint is a family of survey methods for estimating how parts of an offer contribute to preference. Respondents choose, rate, or rank product profiles. A statistical model estimates the contribution of each feature level, called its part-worth utility.

What does a conjoint question look like?

A choice task might ask someone to pick between these hypothetical plans:

Neither plan wins on every dimension. Across repeated tasks, choices reveal how respondents trade storage and support against price and commitment.

What are attributes, levels, profiles, and utilities?

Utilities have a model-dependent scale. Compare them within the study; they are not universal units of customer value.

What are the main types of conjoint?

Choice-based conjoint asks people to select a profile, sometimes with a “none” option. Adaptive conjoint changes later tasks based on earlier answers. The design should match the decision, number of attributes, and respondent effort.

How do you design a conjoint study?

1. Name one decision, such as choosing a package or pricing a feature.
2. Use clear attributes and realistic levels that respondents can compare.
3. Vary the levels enough to estimate their effects. Avoid impossible combinations or an offer that wins on every attribute.
4. Consider a “none” option when refusing all offers is realistic.
5. Plan sample size for the model, tasks, precision, and customer groups you need. Pilot before launch.

How should you interpret conjoint results?

• Utilities: Compare relative preferences and their uncertainty.
• Attribute importance: Depends on the range of levels tested. A wider price range can make price appear more important.
• Willingness to pay: Requires an appropriate price model; weak or nonlinear price effects can make estimates unstable.
• Choice simulations: Estimate preference among the offers entered. Awareness, availability, and real budgets affect actual market share.

When should you use conjoint analysis?

Use it when you can describe realistic offers and need to compare features, packaging, and price together. For a simple item ranking, use MaxDiff. For satisfaction categories, use Kano. For perceived price boundaries, use Van Westendorp.

What makes conjoint results unreliable?

Dense profiles encourage shortcuts. Vague levels leave people comparing different imagined products. Respondents outside the target market can give precise answers to the wrong question. Inspect model fit and uncertainty, then validate the chosen offer with observed behavior.

How do you use conjoint in SenseFolks?

Create a UserChoice survey and define the attributes and levels. Preview the tasks with intended respondents, then embed the survey where they can evaluate the offer. Read the results with your design assumptions beside them.

Frequently asked questions