That's a question coming out of the debate set off by Tim Bray's "Doing It Wrong" comparison of big IT projects with big Web 2.0 efforts such as Facebook and Twitter. The degree of waste engendered in the status quo of enterprise IT operations and development is shameful, Bray says, and it's time for lessons to be learned. One of these is that the firm commitment to the use of third-party utility computing resources and the deep assimilation of core elements of agile development methodologies is a significant contributor to Web 2.0 success, and that enterprises would do well to adapt.

As is often the case with such posts, the most interesting things are coming out in the comments. Commenters point out that the objectives and orientation of the efforts are considerably different, and that the enterprise environment is a far cry from the loosely coupled startup environments in which the Web 2.0 hits began (although I think that this is, at least in part, Bray's original point). They also point out that apart from the mega-successes, there are a lot of failures in the Web 2.0 world, too; clearly the methodologies in and of themselves are no guaranty.

As I argue elsewhere, I think the real key is that agile adoption can make failure cheap. It doesn't necessarily reduce failure rates, or at least that is my sense of things (I'm not aware of any studies in the matter; if anyone else is, I'd appreciate a pointer to them). It does reduce the costs to the extent that failures need not be the painful, over-the-top spectacles that Bray cites (and that most of us are well familiar with). It does this not by being magically more effective, but by reducing scope and moving the goalposts closer to the kicker. And so the real question, to me, is whether or not enterprise leadership is capable of defining project goals in a manner where they might be successfully addressed by agile efforts.