How should I use Conjoint Analysis?
Some of the main applications for Conjoint Analysis are: testing the appeal of a new product, understanding product deletions, portfolio optimization, product optimization, assessing the impact of changes in product design, pricing optimization, understanding the psychology of the buyer from purchase hierarchies to different preferences, computing brand equity and market segmentation. Check out “The Main Applications of Conjoint Analysis” for more information!
Why do Conjoint Analysis?
Choice-based conjoint analysis is the closest thing we have to a crystal ball. It allows us to answer ‘what if’ questions about things that have not happened yet. Pretty cool, right? It does so in a way that explains how differences in products and attribute levels contribute to the appeal of the products.
A Conjoint Analysis Example
Let’s say for example, your company produces chocolate and wants to understand the preferences for different brands (Lindt, Godiva, Hersheys) with three different types (white, milk and dark chocolate) at three different price points ($2, $3.50, $5). There are several ways to perform a conjoint analysis experiment here.
Conjoint Analysis Experimental Designs
The easiest way to create an experimental design is to randomly choose brands, types and price points. However, if you use this random approach, you may run into a problem with balance. There are ways to try and solve this like increasing the number of versions shown or using a design that ensures balance. You can also increase the accuracy and efficiency of your design by avoiding ‘easy’ questions and having prohibitions which are rules that stipulate that certain combinations cannot be shown.
You could also create an efficient design with priors, a labeled choice experiment or an alternative-specific design. If you would like seven or more attributes (brand, type, price etc.) we recommend using a partial profile design. Check out “Experimental Design for Conjoint Analysis” for more information!
Conjoint Analysis and Data Visualization
Don’t limit yourself to just a simulator or a few pie and bar charts to show the results of your conjoint analysis. Go further with utilities plots, small multiples of histograms, correlation heatmaps, substitution maps and indifference curves.