Conjoint Analysis

Preparing the experiment

To perform a conjoint analysis, first narrow down the top 3-5 product attributes you want to test. Choosing which attributes to test and in what priority should be qualified by generative research such as Customer Discovery Interviews as a preparation to the experiment.

The more attributes you choose to test at the same time, the more responses are needed. As a general rule of thumb, the appropriate sample size can be calculated by:

Sample size = (Total count of tested attribute values – Attribute count) · 10

By mixing all possible variations of the various attributes and its values, you will generate all product attribute variations. There are several software tools that will help you do this.

Example case

To illustrate conjoint analysis in action, consider the following example examining what kind of phone plan would sell the best (source):

Attributes	Possible attribute values
Brand	Brand A Brand B Brand C Brand D
Price	$60/month $75/month $100/month
Minutes included	800 1,000 1,400 2,000
Rollover options	No rollover of unused minutes Unused minutes rollover for 1 month Unused minutes rollover for 1 year
Call options	No free calling based on contacts Free calling to top 5 contacts Free calling to top 10 contacts

A limited number of available product variations are chosen to be presented to participants side by side. One combination of product variations for participants to choose from could look like this:

Brand A	Brand B	Brand C
1,400 minutes	1,000 minutes	800 minutes
Unused minutes rollover for 1 month	No rollover of unused minutes	Unused minutes rollover for 1 year
No free calling based on contacts	Free calling to top 5 contacts	Free calling to top 10 contacts
Costs $100/month	Costs $75/month	Costs $60/month

To dertermine which attribute values are more powerful choice indicators, how often a product value was included in the product selected are counted and presented as the percentage of times the attribute value was included in the selected product.

For the Call options attribute, results could look like this:

Call Options	Value
Free calling to top 10 contacts	50%
Free calling to top 5 contacts	20%
No free calling based on contacts	0%

For the Rollover options, results could look like this:

Rollover Options	Value
Unused minutes rollover for 1 year	100%
Unused minutes rollover for 1 year	30%
No rollover of unused minutes	0%

In the example above, the Call Options had values ranging from 0 to 50 and the Rollover Options had values ranging from 0 to 100. This would indicate the Rollover Options was more important to participants than the Call Options.

Primarily used for existing markets

Conjoint analysis concerns itself about uncovering customer preferences (not actions). This is why using this methods primarily makes sense in existing markets, where product attributes are generally known by participants.

In the early stages where totally new attributes are introduced to participants, it can be hard for participants to comprehend and understand what they are choosing between. This is when results become inaccurate with the risk of producing false negatives.

This is why it makes sense to conduct significant explorative and generative research before starting a Conjoint Analysis experiment. This method is rarely used in early stage innovation products, but after product/market fit has been reached.

Two types of Conjoint analysis

In market research, two prominent methods of conjoint analysis, Discrete Choice-Based Conjoint (CBC) Analysis and Adaptive Conjoint Analysis (ACA), offer unique insights into consumer preferences and decision-making processes.

Discrete Choice-Based Conjoint (CBC) Analysis is a method where respondents are presented with several product or service options, each with varying attributes, and are asked to choose their preference. This approach mimics real-world buying scenarios, where consumers make trade-offs between different attributes like price, quality, brand, and features. CBC is particularly effective in understanding broad market trends, assessing price sensitivity, and studying how different product features interact to influence consumer choice. This methodology is advantageous when the decision-making scenario involves a limited set of key attributes and where understanding general market preferences is crucial.

Adaptive Conjoint Analysis (ACA) offers a more personalized survey experience. Unlike CBC, where all respondents see the same set of options, ACA adapts the questions based on the respondent’s previous answers. This method is suited for scenarios where products or services have a large number of attributes, and detailed insights into individual consumer preferences are crucial. ACA is especially useful for custom product development, where understanding the nuanced preferences of each consumer can guide personalized product or service offerings.

The choice between CBC and ACA largely depends on the specific objectives of the market research. CBC is the go-to method when the focus is on understanding general market trends, price sensitivity, and preference trade-offs between a few key attributes. It’s suitable for studies aimed at optimizing product features or pricing strategies for a broader market. In contrast, ACA is more appropriate when the research requires in-depth insights into individual consumer preferences, particularly when dealing with a large number of product attributes. It is ideal for businesses focusing on customized products or services where individual preferences significantly influence the product development strategy.

Use cases for Conjoin Analysis

This method is useful in several situations.

Feature Prioritization

When developing or updating products, deciding which features to include can be challenging. Conjoint analysis helps in prioritizing these features based on how much they contribute to the overall value perceived by the customers. This approach ensures that the most valuable features from the consumer’s perspective are included, enhancing product appeal and customer satisfaction.

Pricing Strategy

Conjoint analysis is highly effective for pricing optimization. Companies can gauge how price changes impact consumer choice and determine the optimal price point for their products or services. By including price as one of the attributes in the study, businesses can understand the price elasticity of their products and identify the price range that maximizes both sales volume and profit margins.

Market Segmentation

Conjoint analysis aids in market segmentation by revealing different preference patterns among various consumer groups. By understanding these patterns, businesses can segment their market more effectively, tailoring their products, marketing efforts, and services to meet the distinct needs of each segment. This targeted approach can lead to more efficient use of marketing resources and higher customer satisfaction.

Competitive Analysis

In competitive markets, understanding how your products or services stack up against competitors is vital. Conjoint analysis can simulate competitive market scenarios, helping businesses understand their competitive position. This method can reveal how changes in product features or pricing might shift consumer preferences between different brands, providing valuable insights for strategic planning.

Sales and Distribution Strategy

Conjoint analysis can also inform sales and distribution strategies. By understanding consumer preferences, companies can make informed decisions about which products to stock in which locations, how to display products in stores, and which channels (online vs. brick-and-mortar) might be most effective for different products.

Customer Needs Analysis

This tool is valuable for digging deeper into customer needs and desires. It can unearth latent needs that consumers themselves might not be explicitly aware of, providing an opportunity for innovation and differentiation in the marketplace.

Risk Reduction in New Product Launches

Launching a new product is inherently risky. Conjoint analysis reduces this risk by providing empirical data on how likely a product is to succeed in the market. This data-driven approach can inform go/no-go decisions and guide investment in new product development.