If you read one thing today, read The Fishing Maturity Model post. While I have seen some fantastic process improvements, I've also seen some pretty bad ones. Personally, the best process improvement I've seen has been moving towards agile processes and collaborative testing and development under any name, but any way where you are vetting ideas together, reviewing eachother's code and planning, and improving and sharing knowledge together. For me the worst is when every little thing is scrutinized to the nth degree, over documented, and stingy experiments begin trying to save money without proof they will pay off but still they harm the software and the people. I call this Malevolent Tampering privately and wish for it to stop, but try to take a "we'll see" stance which means I consider it unknown and unproven if it will work well in the context it is being tried and would not make bets one way or the other. Anything which reminds you too much of "The Bobs" from Office Space? Beware the Fishing Maturity Model and start asking questions.