Current practices in software testing such as Test Driven Development or Behavior Driven Development aim at linking code to expected behavior. In this context, code coverage is widely used to improve code quality, reduce bugs and ssure requirements satisfaction. Even if change tracking software allows finely analyzing code evolution, associating a particular code chunk to the requirements at the origin of the code modification is difficult for a large code base. In this preliminary work, we propose a new "user story coverage" metric that reports lacking requirement coverage quality, to help developers focus their efforts on enhancing unit and integration tests. We propose a methodology to compute this metric in a robust and automated fashion and evaluate its feasibility on open-source projects.