Product Optimization and the Acceptor Set Size

The acceptability of a product, measured by the acceptor set size (percentage of consumers rating the product acceptable), is a function of the perception of its attributes. The attributes are themselves a function of the inputs to the product (such as ingredients, processing or storage variables). These assumptions lead to the following model: Acceptor set size = F (attribute 1 , attribute 2 , …, attribute n ) Attribute j = f (input 1 , input 2 , …, input m ) If we assume that these functions are differentiable, we can estimate the partial derivatives of the acceptor set size, with respect to the input variables. The gradient vector obtained indicates the fastest way to maximize the acceptor set size. The gradient search method, using the acceptor set size as an objective measure, can be applied in a variety of situations: to improve existing products, to maximize the acceptability of new products, and to study the relationship between shelf‐life and acceptability.