Predicting Commercial Success of Products Based on Product Testing Data
The Challenge
A major Consumer Packaged Goods (CPG) company was challenged with determining where to focus its commercial resources based on data collected from product testing. The product testing data include information on the products and how testers rated the products.
The client desired to utilize its testing data to make future predictions about the commercial success of the products and to learn which product attributes to optimize in order to maximize commercial success of the products. To achieve these ends, the client engaged the Dataspora team.
The Solution
Dataspora leveraged the REFS™ technology platform to decipher the underlying relationships between product attributes and commercial success.. These results were used to build predictive models of commercial success. The models identified combinations of product test attributes that may influence commercial success and potential areas for improvement in the product tests.
The product testing data included dozens of product- and study-specific variables. Dataspora used its REFS™ platform to analyze the survey data and create probabilistic models. REFS™ allows one to learn causal models in an unbiased, systematic and automated fashion. Ultimately, Dataspora produced a market performance prediction tool that could be used by the CPG company to prospectively predict commercial success.
Figure 1. Simulated predictions of commercial success based on product testing data. Product Attribute 1, 2 and 4 should be main focuses for R&D spend, since improving customer satisfaction for those leads to significant commercial success. Product Attribute 3, however, only requires a ‘medium’ satisfaction rating to maintain success.
REFS™ can produce models that are not only predictive, but identify probabilistic cause and effect relationships. This latter feature allows the client to ask questions to inform future actions by understanding how changes will impact outcomes of interest. For example, the client can ask “Which product attribute has the greatest effect on commercial success?”, “Which product attribute should receive the most R&D spending such that we maximize its potential for commercial success?” (Figure 1), “Which product attributes only need to be ‘good enough’ such that unnecessarily high additional budgets are avoided since they will not have major effects on commercial success?” and “What non-obvious combinations of product features have a measureable impact on commercial success?”
Additional example conclusions resulting from the collaboration include:
- Product Attribute A has the most linear relationship with commercial success and is also the strongest driving force.
- Product Attribute B isn’t as important as Product Attribute A but has a greater effect on commercial success than Product Attribute C.
- Product Attribute C and Product Attribute D have the least effect on driving commercial success.
- Product Attribute C only has to be “good enough” as there is not much change in commercial success when improving the product attribute rating from “fair” to “excellent.”
The Benefit
The CPG company received invaluable information from the models built by Dataspora, including specific recommendations on how to use their resources to maximize commercial success. This information also allows the client to optimally allocate their budget, so that only the product features that drive revenue more significantly receive more resources than features that are weaker drivers of commercial success.
Finally, the CPG company received feedback from Dataspora regarding which data sources would be most valuable in improving the model and its predictions. The initial phase of this collaboration with Dataspora provided the client with valuable models predicting commercial success from product attributes. The next step for the client is to gather the additional data required to build even more comprehensive models and to integrate them into their operational workflow. Using the insights provided by the Dataspora models, the client will be able to focus its resources such that every one of its products is given the best possible chance to reach maximal commercial success.
