In this research, the product design optimization is addressed. The objective is to maximize an objective value such as margin, revenue or market share by composing a product based on multiple attributes. The Nested Partition (NP) method in com- bination with Genetic Algorithms (GA) and a machine learning method (CTREE) is used to optimize margin for a hairdryer manufacturer. Methods lead to high increases in margin compared to the current scenario in the hairdryer market.