> ... IT software development profession largely lacks — or fails to put into place — automated, objective and repeatable metrics that can measure progress and

> ... IT software development profession largely lacks — or fails to put into place — automated, objective and repeatable metrics that can measure progress and predict project completion with any reasonable degree of accuracy. Instead, we tend to rely on seat-of-the-pants (or, less politely, out-of-one’s-butt) estimations by IT engineers or managers that a given subsystem or application is “80% done”... Here was where I started to realize that this was going to be a bit of a hand-wavy ("out-of-one's-butt") argument. Are we talking about tracking progress or estimating? This seems to conflate the two. > Second, IT engineers by nature tend to be optimists, ... Really? Optimists? I don't think that word means what you think it means. This "optimism" wouldn't have anything to do with being pressured by management to give more favorable estimates, would it? What makes both estimation and progress tracking difficult or impossible is a lack of good requirements specifications or use cases. I have never worked for a company where anyone took these seriously. Sure, they talked a good line, but never seemed to follow through. There's always an excuse. The thing is, if you don't know what you want and/or can't express it, you're unlikely to get what you want.