Six Sigma is often thought of as a manufacturing discipline, but lately, I've seen it popping up more and more in IT. There was a great story in Computerworld today about how Bank of America is using Six Sigma to improve their development methodology.
While people often think of Six Sigma as a bunch of statistics geeks running around pretending that they are martial artists ("I'm a green belt". "Oh, yeah? I'm a Black Belt."), the real key to Six Sigma is that it helps you create repeatable processes. I know that the idea of repeatable development doesn't square well with the "software development as art" idea, but it really is key to long term quality.
And, more importantly, it will become even more key over the next 20-30 years as we try to deal with the loss of talent. This is how BofA is really thinking about it. From the article:
"Desoer is optimistic that the standardized methodology will reduce development time and help make developers and project team members more transferable across business units in coming years. That could be crucial if an exodus of retiring baby boomer technologists and a widely anticipated shortage of entry-level IT workers make it tougher for the bank to find and recruit people with the skills it needs, Desoer says. â€œWe are going to be increasingly challenged to find highly qualified technology associates five to 10 years out,â€ she adds."
There are three ways to deal with the increasing War for Talent that is going on out there. The first approach is to do what most people do - accept that your talent is going to be mediocre, and pretend that it's not. The second most common approach is to work harder and harder to recruit more and more - spending like a drunken sailor on recruiters fees, internal recruiters, job boards, evaluation tests, etc. This one's actually what the really great companies out there right now are doing. And, if you do it right (like we did when we started the Toronto office at nCircle), you can come up with incredible talent.
Of course, that strategy only works as long as most people are employing the first approach (the Ostrich method). In the long term, it will be the companies that take on the approach that B of A is using that will be truly successful - the ones that combine the intense hiring and talent screening of the second approach with a long term plan that focuses on repeatable process to ensure a minimal level of quality (whether quality means "time to market", "bug free", "feature rich", etc.) that will truly win the game.