Conjoint Analysis. What is it and when to use it?

Good, beautiful, and cheap. Humans always want products that have it all! But in real life, when it comes to making a purchase decision, we are often willing to sacrifice some non-essential features to enhance others. Conjoint analysis is a type of market research study that helps companies identify which combinations of features will have the greatest acceptance to ensure the success of innovation. Are you ready to learn more about conjoint studies?

What is Conjoint Analysis?

A conjoint study, conjoint survey, or conjoint analysis is a quantitative research technique used to identify combinations of product features that are more likely to be chosen when making a purchase.

When to Use Conjoint Analysis?

Conjoint analysis has various applications:

• Choosing winning feature combinations. When launching a new product, you need to decide which combination of features to offer to customers. Conjoint analysis helps identify combinations that will be more accepted in the market. This allows companies to focus development efforts on those with the most potential.
• Understanding the value consumers place on each feature. When a feature consistently appears among the preferred options in the conjoint study, it indicates its importance. A company can then focus its design and communication efforts to emphasize it over competitors.
• Predicting the success and market share of a new launch. Conjoint analysis can show to what extent the new product will be chosen over existing options in the market. If you have information on the market share of these products, you can estimate how the new launch would perform.
• Identifying the most profitable products. By combining cost information with preference levels and price, you can select the most profitable products to include in the portfolio.
• Analyzing which combinations certain consumer profiles would prefer. This way, you can create a range of winning products specific to each consumer profile.

Attributes and Levels of a Product

So far, we’ve been talking about product features. Let’s be more precise in the terminology and talk about attributes and levels. Attributes are each of the product’s characteristics. For example, when purchasing a new mobile phone, the phone can have different attributes: brand, screen size, storage capacity, chip power, or price. And each of these attributes has different levels. For example, storage capacity may be measured in GB, and you can find phones with 64 or 128 GB. For another attribute like the brand, you can find different «levels» like iPhone, Samsung, or Xiaomi, to name a few. Thus, each product has different attributes, and each attribute has several levels.

This example can be extended to any other sector. For instance, when choosing a dairy dessert from the supermarket shelf, the consumer will consider attributes such as taste, quantity, nutritional benefits, brand, or price.

Utility Concept in Conjoint Analysis

Consumers buy a product because the utility it provides, the benefits of having that product, are greater than the money they pay for it. When choosing a product, consumers try to maximize utility for the price paid. To do this, consumers, more or less rationally, evaluate attributes and their levels to make a decision that maximizes the utility they receive. Conjoint analysis is based on utility theory to force respondents to define their preferences.

Types of Conjoint Analysis

There are various types of conjoint studies:

• Conjoint Value Analysis (CVA). Respondents are asked to rate each combination of attributes and levels on a rating scale, such as a numeric scale from 1 to 10. Preferred combinations will receive higher ratings, making it relatively easy to calculate the weight of each attribute and level in respondents’ decisions.
• Adaptive Conjoint Analysis (ACA or adaptive conjoint). Similar to CVA in that respondents are asked to rate each combination of attributes and levels. The difference is that the combinations presented take into account previous responses. This limits the number of options presented to the consumer, reducing the interview duration and fieldwork cost.
• Choice-based Conjoint (CBC). Instead of rating each combination, respondents are presented with various sets of combinations and asked to choose which one they would buy from each set. This method simulates a decision much closer to real-life purchasing decisions. CBC involves a more complex mathematical calculation, but its similarity to actual buying decisions makes it the preferred method in most cases. The adaptive version of this conjoint is called Adaptive Choice-based Conjoint.

Conjoint Analysis vs. MaxDiff

MaxDiff analysis is another technique very similar to conjoint analysis. The difference is that in MaxDiff, respondents are not asked to choose which combination of attributes and levels they would buy but are instead presented with several levels of the same attribute and asked to pick their most and least preferred options. This method evaluates preferences more effectively than a Likert scale or other rating scales.

The reason MaxDiff is often included among conjoint methods is that the statistical analysis used is very similar to that used in Choice-based Conjoint.

How to Conduct a Conjoint Analysis?

Conducting a conjoint study involves several phases:

• Define attributes and their levels. The first step is to decide which attributes should be considered and what levels make sense to offer to consumers. It’s important to note that there may be impossible or undesirable combinations that need to be eliminated. For example, greater product quantity and quality may require a higher price. Therefore, restrictions are typically introduced to avoid options that consumers would clearly choose or discard. These restrictions also include options that are impossible for brands to offer or economically unviable.

• Conduct a survey. Using a questionnaire, different options are presented along with visual aids to allow consumers to simulate a shopping situation.

• Perform statistical analysis. A conjoint survey requires sophisticated analysis that goes beyond the usual frequency charts provided by research platforms. Therefore, data should be analyzed by experts who are well-versed in using statistical analysis software.

At We are testers, we have an advanced service team that can assist you throughout the entire process of creating your conjoint study, including the analysis and recommendation phase. If you’re considering conducting a conjoint study, contact us, and we’ll provide you with all the information you need.