First separate out actual root causes. I prefer the "Committed to Completion" ratio because it's a metric that makes no sense to game. It's simply, ask people

First separate out actual root causes. I prefer the "Committed to Completion" ratio because it's a metric that makes no sense to game. It's simply, ask people to commit to units of work they will complete by a certain date. Then measure how many of those units were completed by that date. From there you can separate out over-committers, under-committers, effective contributors, and slackers (people who intentionally commit to very little). Then you can more narrowly discuss with the various cases what is leading to that outcome. The work itself also needs to be interesting if you want self-motivation. A skilled team with proper support and tools managed lightly, with you removing any blockers in their way, whose work is interesting and who are shielded from organizational politics will often perform well without needing to push.