Only Better Practices.
The phrase "Best Practices" came up in a meeting today, and I had forgotten my epi-pen. It was touch-and-go for a few minutes.
Anytime someone suggests that you follow best practices, you can be fairly sure they are trying to sell you some crappy software or a methodology you don't need. Alternatively, they might not be confident enough to make a judgement for themselves and prefer to rely on a crutch.
I'll give an example: several years back a famous IT consulting firm (left unnamed because I'm feeling charitable) published a white-paper stating that the optimal server to sysadmin ratio was 10-12. At that time my team's goal was 300. Which practice is best? It depends on context. In a small shop with bespoke servers, their numbers might fit. Our context was supporting a large-scale application which was architected to require minimal hands-on administration. Your practices should fit your context and not a framework or consultant's recommendations.
Methodologies sell books, consulting gigs, and training contracts. Agile, ITIL, CMMI, the list goes on and on. They are not entirely without value, but adopted blindly they become constrictive. You can meticulously log change tickets for a workflow to be ITIL compliant or you can spend some cycles automating the workflow. Which is the better practice?
People and companies on the leading edge don't follow best practices. They create their own. If Google had followed the best practices of the time when they started, they would have bought giant Unix boxes, stored their indices on enterprise SAN, and gone broke.
Look at the methodologies. Look at what other companies do. Look at what other industries do and find analogies to yours. Cherry pick. Use your own judgment. If it seems like a Better Practice than yours, try it.
Maybe if you do it right you can slap a name on your methodology and sell some books.
It seems appropriate to let Scott Adams have the last word here.
