Conjoint research is a statistical means of knowing what customers like and what influences them to make the decisions on the said products. The aim of this study is to be able to model customer products and services to either create demand or to conquer competitive markets.

To understand the principals behind conjoint analysis, we must first break a service or product to the basic parts it’s composed of, called attributes and levels. These parts are the ones tested to evaluate the customer preferences. These parts are then combined together using statistical designs and techniques in a way that they can be easily tested to determine the customers’ preference.

A well designed study using experimental designs is capable of using statistical analysis to calculate the value or utility score of every part of the service or product so that the reason why the customer made the decision is finally arrived at.

The simplest example often used to illustrate conjoint research is in TV marketing. The attributes that may be assigned to a TV include the display size, smart technologies, and brand.  Each of the attributes is then subdivided into levels. The levels for display size for instance can be 23 inch, 32 inch, 40, inch and 50 inch. After listing different attributes and levels for different products, you can then use it to create a set of product profiles.

A product profile is the most possible combination of levels and attributes. This helps to create a set of choices (choice sets) from where the customer or respondent can pick from. An increase in the number of attributes increases the choice sets. Different techniques are involved in determining the numbers and types of choice sets and the number of profiles according to the type of research, so that the most amount of information can be collected.

Types of conjoint analysis include choice based conjoint (CBC), Adaptive Conjoint Analysis (ACA), Adaptive Choice Based Conjoint ((ACBC) and others that determine the complexity levels that is required for specific research to be fruitful.

After respondents have completed the choice section, their decisions are analyzed using statistical tools that produce implicit numerical evaluation for every attribute. They are then modeled to project the preferred choices by the majority of respondents.