This ones tough. In my life I do 2 things:<p>1) Logging. Measure what you want to improve, and then actually track the measurements. Also: log interesting things - your git repo tells all kinds of stories, if you want to put the effort in - because it's tracking something through time. Other forms of logging can also be used retrospectively when you didn't have a goal at the time of logging.<p>2) Ask for outside opinions (and log them). These are both better and worse than anything you can do yourself. Others have a much better perspective than you on the issues, but there is rarely an incentive to be honest in "friendly reviews".<p>These are both pretty obvious. And not silver bullets, but I do think they work particularly over longer time periods.