Conjoint Analysis has evolved to be one of the most powerful analytic
tools in research, with far reaching tactical and strategic uses.

Predictive analysis, discrete choice conjoint, price elasticity - these are the many aliases for what is commonly known as trade-off analysis.  Before the invention of microprocessors for desktop computers, choice conjoints would demand floor-to-ceiling of computing hardware, months of data crunching, and large teams of data scientists.  Furthermore, technology allows us to simulate real life shopping trips on web-based surveys using virtual reality.  In the conjoint design, respondents are asked to make a purchase selection from a series of competing mock-ups with varying attributes e.g. brand, price, speed, material, etc.  Econometric modelling allows for the estimation utility functions, which ultimately gives insight into how sensitive customers are to slight changes in product features.  The strategic implications can be profound, guiding pricing scenarios, advertising expenditures and market share.  Collaborative workshops with strategists and cross-functional teams are key to leveraging insights.

Traditionally, price elasticity studies have been profoundly complex and expensive.  But innovation and technology is changing that.  Survey platform providers like SAP (formerly Qualtrics) have now developed their own discrete choice conjoint platforms that have made building "market share" simulators as easy as a click-of-a-button.  But just as Factor Analysis and Discriminant Analysis can be processed in seconds instead of by a group of mathematicians over several months, survey instrument design and target audience is key.  As powerful as these tools are, poor design often leads to results that are neither practical nor actionable. 


Below is an example of a Choice Based Conjoint (CBC) where participants are shown 3 Smartphones with 6 features. In this choice exercise, participants are asked to choose from one of the four options.  This is repeated 12-18 times (depending on the complexity of the study), where differing smartphones and their features are rotated in.  Results from a CBC would allow us to...


+ understand the relative importance of one feature over the other e.g. camera features are more important than operating systems

+ predict the hypothetical market share that would result if Amazon were to lower its fire phone price or introduce a 16 Megapixel camera

+ understand the price elasticity for the brand and its associated product features

One important thing to keep in mind is that marketing campaigns, advertising, how competitors react to price reductions are dynamic, so though CBC can give you a hypothetical prediction of market share, it is a snapshot in time that assumes all else are equal in the marketplace.

Features below can also vary to include shipping costs and retailer (eg vs vs  In the design, it's crucial to carefully consider the attributes that go into the collective decision-making of the consumer.      


Food Case Study

A consumer-packaged goods company is planning to launch a new variation of a cereal product in U.S. stores and would like to understand pricing scenarios and impact on market share.  Many characteristics of the product must be considered, including brand equity, cannibalization, advertising, flavor, nutrients, size and existing competitors. In order to understand how consumers make trade-offs in product attributes in their decision-making process, sophisticated tools made viable by the advancement of supercomputers has brought forth discrete choice modelling. An experimental design is created to capture all relevant attributes and shopping scenerios are presented to customers. Statistical rigor will allow researchers to infer the responses from several hundred respondents to the greater market, resulting in a forecasting simulator that allows clients to adjust certain inputs to predict revenue and profit from the new product launch. Action planning workshops can be leveraged to formulate strategies around pricing and product positioning.


© 2011 by LP