Choice-Based Conjoint (CBC) Analysis

Introduction to conjoint analysis

Conjoint analysis is a category of research methods, among which choice-based conjoint is the most popular, that mimics the respondent’s real world tradeoffs when making decisions. It is used for pricing studies, product optimization, healthcare options and many other things. To get a more in-depth understanding of conjoint analysis, refer to our page on conjoint analysis.

What is choice-based conjoint (CBC)?

When you want to understand and predict how people make choices when facing challenging tradeoffs, Choice-Based Conjoint (CBC) is the most widely-used survey-based approach. CBC also is known as discrete choice modeling (DCM) or discrete choice experiments (DCE).

Related video: Creating a Choice-Based Conjoint Exercise with Discover

What is the difference between conjoint analysis and discrete choice?

Choosing a preferred product from a group of products is a simple and natural task that everyone can understand. The difference between choice-based conjoint and the earliest approaches to conjoint analysis is that the respondent expresses preferences by choosing from sets of concepts, rather than rating or ranking them.

How Does The Choice Model Work?

In this example of a choice model, the user is given the choice of four truck types with various attributes.

Example of a choice task

The combination of attribute (feature) levels we ask respondents to evaluate is critical to making choice-based conjoint analysis work properly. Fortunately, Sawtooth Software takes care of those details (though you may import your own designs if you wish).

Each level appears nearly an equal number of times and appears with levels from other attributes nearly an equal number of times. This makes for a fair and balanced (orthogonal) experiment where the utility value (the preference) of each attribute level can be measured independently and with high precision.

Build a preference model

We can assess the relative impact of each attribute level on choice just by counting "wins." But, it’s more precise to fit statistical models such as multinomial logit (MNL), latent class MNL, and hierarchical Bayesian (HB) estimation.

HB is the most popular approach and leads to a set of utility scores for the attribute list for each respondent. Latent class can find groups of respondents who are very similar to one another in their choice preferences, while being very different between the groups. Thus, latent class is an excellent approach to leverage choice-based conjoint data for needs-based segmentation and strategy.

Use Cases for Discrete Choice Analysis

Researchers often use discrete choice modeling to study the relationship between price and demand, especially when the price-demand relationship can differ from brand to brand. One of the strengths of choice-based conjoint is its ability to deal with interactions, such as between brand and price, or when different colors work better with different styles.

Choice-Based Conjoint (CBC) is used in marketing and economics applications across a variety of cases and industries, including:

• New product design, existing product redesign or line extension
• Pricing studies
• Market segmentation
• Healthcare choices, including cancer regimens
• E-commerce
• Employee research (health plans, benefits)
• Transportation choice
• Legal disputes
• Green energy, electric vehicles, ride sharing
• Public health
• Education
• Environmental impact
• Finance, banking
• Technology and innovation