Having an effective product testing programme is key to developing products that delight consumers, but all too often the results get forgotten about once the product is launched. Whether it’s sensory or consumer data, if you’ve got a lot of data from product tests you’ve completed then we can harness the power of that data by building it into a database and using advanced analytics, such as Bayesian Network Analysis, to generate strategic insight from it.
Without going into too much detail about how Bayesian Network Analysis works and the stats it’s based on (if you want to know more contact our stats gurus), it is a graphical network model that helps you to see how lots of different variables (in this case product attributes) are related to each other.
For example we can look at how all product attributes are linked to a key variable such as overall opinion, so we know what the most important elements of a product are. These are the ones to get right to delight your customers! In the example above we were able to tell our client that aftertaste was the key thing to get right to improve overall opinion of their product.
Going further, we can tell you which product attributes are linked together, so in the network above we can see that if we change the naturalness of the aftertaste this will impact on the strength of the aftertaste, and that both of these together will impact on overall opinion.
MAKING PREDICTIONS THROUGH BAYESIAN NETWORK ANALYSIS
All of this means that the network will also be able to predict the impact of changing a product attribute on other attributes, including overall opinion, very useful information for product developers!
As if that wasn’t enough, the network will also help us to devise action standards for each product attribute based on the information from your database. For example we can tell you what score you would need to get in a product test to be classified as ‘excellent’.
Get in touch if you’d like to find out more!Subscribe