Auto-evaluation is work. More work. No matter what anyone will say. Someone has to prepare it, and then execute it. And unless this evaluation is done for the sake of it, it entails more work. If the evaluation states that all is as good as can be, this knowledge still has to be made useful. It the evaluation suggests adding a project, obviously, more work. And even if the evaluation suggests removing a project, that removal still is a piece of work.
This is by no means me saying you shouldn’t auto-evaluate or change – this is me saying: beware. As most auto-evaluations are decided once workload gets too big, you must accept that for a certain period of time, it will get worse before it gets better.