David Boswell has a couple of interesting posts (here and here) about how he is using metrics to measure how effective Mozilla is at attracting and engaging people who express an interest in helping contribute to the Mozilla mission.
Some of the metrics being used can be seen at Mozilla's Are We Growing Yet website. What I think is important is how the team is trying to look at metrics from the site to see if tweaks have an impact on attracting more people. Obviously, this is something commercial websites have been doing for a long time and it is about time these lessons were being applied in open source community management.
What I'm particularly excited about (and have helped with) is the creation of contribution paths for different parts of Mozilla. This way we can begin to try to measure how many people that express an interest in contributing to Mozilla start to contribute code (or engage in another activity, like marketing) and how far along the code contribution path they travel. Obviously we'd love for contributors to travel as far along that path as possible and so, with these metrics, we can ideally begin to see how changes in policy, tools or websites might encourage or discourage people's decision to contribute.