Conjoint Analysis

Conjoint analysis is a preference measurement method that estimates how much each attribute level of a product or policy contributes to respondents' choices or ratings of experimentally varied profiles.

컨조인트 분석은 실험적으로 변화시킨 제품·정책 프로필에 대한 응답자의 선택이나 평가에 각 속성 수준이 얼마나 기여하는지를 추정하는 선호 측정 방법이다.

How do I find out which product features matter most to customers?

고객에게 어떤 제품 특성이 가장 중요한지 어떻게 알아내나요?

Conjoint analysis shows respondents profiles that vary systematically on attributes such as price, brand and features, then estimates part-worth utilities from their ratings or choices. It fits product design, pricing and policy preference studies, and choice-based conjoint is the most common current form.

When to use Conjoint Analysis

• Ratings or choices on profiles built from an experimental design of attributes
• Estimating part-worth utilities and attribute importance
• Simulating market shares or willingness to pay

When to use something else

• Ranking a long list of single items rather than combinations; use MaxDiff (best-worst scaling)
• Measuring price sensitivity alone; use Van Westendorp or Gabor-Granger

Assumptions and requirements

• Attributes and levels relevant and clearly defined
• Orthogonal or efficient experimental design
• Compensatory decision rules
• Sufficient respondents and tasks per respondent

What the analysis reports

• Part-worth utilities for each attribute level
• Relative attribute importance
• Model fit statistics
• Market share or preference simulations
• Willingness to pay estimates

How to report it

Conjoint analysis showed that [attribute] was the most important attribute ([value]% relative importance), with [level] having the highest part-worth utility ([value]).

Related methods

• MaxDiff Analysis
• Discrete Choice Model
• Multinomial Logistic Regression

Run Conjoint Analysis on Quorant