This paper deals with mass customization and the association of the product configuration task with the planning of its production process while trying to minimize cost and cycle time. Our research aims at producing methods and constraint based tools to support this kind of difficult and constrained problem. In some previous works, we have considered an approach that combines interactivity and optimization issues and propose a new specific optimization algorithm, CFB-EA (for constraint filtering based evolutionary algorithm). This article concerns an improvement of the optimization step for large problems. Previous experiments have highlighted that CFB-EA is able to find quickly a good approximation of the Pareto Front. This led us to propose to split the optimization step in two sub-steps. First, a "rough" approximation of the Pareto Front is quickly searched and proposed to the user. Then the user indicates the area of the Pareto Front that he is interested in. The problem is filtered in order to restrain the solution space and a second optimization step is done only on the focused area. The goal of the article is to compare thanks to various experimentations the classical single step optimization with the two sub-steps proposed approach.