Optimization of blended wine quality through maximization of consumer liking

A mixture design was used to optimize blended Vitis vinifera wines (Cabernet Sauvignon, Merlot and Zinfandel) for consumer acceptability. Consumer overall liking values were modeled within an augmented simplex centroid mixture design to determine areas of desirability for each consumer. Two major segments of consumers were identified, with membership mainly dependent upon responses to Zinfandel. Samples were less polarizing when the wines were blended as opposed to remaining as single wines. Three optimized blends were created: for all consumers, segment 1 and segment 2. A subsequent consumer study indicated that optimized blends for each segment were not significantly different from the other wines. However, segment 2 consumers preferred the segment 2 blend to the wine blend optimized across all consumers. The optimized blend across all consumers was the least liked for each segment. Identifying consumer segments can help better optimize products for successful entry into the marketplace. Optimizing a product within a mixture design environment is a useful, efficient technique to understand ingredient interactions and how they affect overall consumer acceptance.

► Optimization within a mixture design environment is a unique approach in sensometrics. ► Clear consumer segmentation dependent upon responses to one particular variable. ► Mixture designs provide ideal product formulation, not only coordinates to determine optimal blend.