Measuring Consumer Relevant Differences: Project Babysteps – Silently Changing Product Formulations Without Losing (Loyal) Consumers

Dr. Danielle van Hout
Science Leader, R&D
Unilever

Signal detection theory & Thurstonian modelling allow the standardized quantification of the size of sensory differences using assumptions of perceptual variability and cognitive processes. They are uniquely valuable to measure the sensory similarity of two or more products.

Their usefulness for industrial research is demonstrated in the case study of project Babysteps, in which a successful product needs to undergo a silent reformulation without losing (loyal) consumers; a frequently occurring FMCG challenge! Discrimination tests with consumers and with a trained sensory panel, and hedonic consumer tests were conducted to identify the size of steps in reformulation that could be taken (each smaller than a consumer relevant size), and to determine how many steps would be needed to reach the specified target product. The relationship between consumer and trained panel discrimination results made the translation in product formulation design factors easy, so that reformulation steps could be accurately defined. By integrating discrimination and consumer hedonic results, potential risks of introducing the reformulated products could be estimated.