Customer satisfaction is a crucial factor in the evaluation of business products, systems and services. Two ways to improve products customer satisfaction have been suggested. The first kind, known as personalized product recommendation system (PPRS), is at the product usage stage. It has received extensive research [3, 4, 5, 6, 7]. The second kind deals with the product design stage. The information of customer needs is collected for product design and improvement. In this paper, we investigated the product design parameters optimization method based on with constrained mixture design and combinatorial fusion analysis. Real world example using the combined method demonstrated that our approach is robust and effective.