LinkedIn Twitter
Well-known expert on why IT projects fail, CEO of Asuret, a Brookline, MA consultancy that uses specialized tools to measure and detect potential vulnerabilities in projects, programs, and initiatives. Also a popular and prolific blogger, writing the IT Project Failures blog for ZDNet.

One response to “Tim Bray’s Enterprise 2.0 conundrum”

  1. Tony Law

    Good posting. But this isn’t new with web development. It was figured out not far off twenty years ago and my comment then was that it was a formalisation of the way I always worked anyway.

    Have a look at the guiding principles of DSDM (Dynamic Systems Development Methodology): small time-boxed steps, user involvement all the way, iteratively refine the deliverable with the users, deliver small stages that work and are useful rather than going for the big bang, recognise that requirements always change over the lifetime of a project because (a) business evolves and (b) using the early deliverables changes users’ ideas.

    Commercial and government organisations have responsibilities (to regulators, customers and suppliers) and do need to have some assurance that there’s a process being followed. Nobody sues Twitter if it fails, but the commercial world doesn’t have that protection. DSDM was created to provide the necessary formalisation in the agile development world.

    Conversely, I attended a meeting a couple of years ago about “What goes wrong with big public sector projects”, convened by the parliamentary IT group, when someone from one of the big developers seriously said that users should sign off the requirement and then go away for three years while the developer builds it … ‘Nuff said, I think.