Measuring productivity in software factories is very important as it seeks that organizations develop greater efficiency and effectiveness in their activities. Productivity is related to the effort required to accomplish the assigned tasks, however there is no standard way to measure it, which makes it difficult to establish policies and strategies to improve the factory. In this paper, we present a model based on Data Envelopment Analysis to evaluate the relative efficiency of the software factories and their projects, focusing the measurement on the Software Production Component. The proposed model consists of two phases, in which the productivity of the software factory is evaluated respectively, and the productivity of the projects it carries out. The model has been applied on 6 software factories with 160 implemented projects, demonstrating that the proposed model allows to determine the software factories and the most efficient projects.