The Load-Bearing Walls of the House of Agile

Posted by Robert Merrill on January 30, 2009 under Agile Methods | 3 Comments to Read

A few weeks ago, I ran across a very provocative post by James Shore called The Decline and Fall of Agile, followed by an equally provocative (and long) string of comments.

Mr. Shore argues that many firms think they are adopting Agile, specifically Scrum, without really understanding what’s essential and what can be safely left out. I heard my first hint of this around the pool at SD West in 2003, with someone saying, “We do XP, except we work 80-hour weeks, scope and schedule are fixed, and we’re not doing any automated testing.” “Here we go again,” I remember thinking.

One of the many commenters pointed out that Agile is merely following the Gartner Hype Cycle, and is now sliding down the backside of the Crest of Inflated Expectations towards the Trough of Disillusionment. That sounds about right to me.

One of the things I tell my clients is that Agile methods are by their nature adaptable and pragmatic, but that they also have the equivalent of load-bearing walls. Destroy the integrity of one of those, and you collapse Read more of this article »

UW-Milwaukee CIO on Project Success

Posted by Robert Merrill on January 27, 2009 under Project Set-Up | Be the First to Comment

In a podcast interview with ZDNet “IT Project Failures” columnist Michael Krigsman,  Bruce Maas, CIO of the University of Wisconsin—Milwaukee sees the roots of success or failure early in project initiation. In his experience, successful project set-up establishes:

  • Relationships
  • Shared vocabulary
  • Shared definition of success

For going on ten years, I’ve specialized in project set-up. It’s part business and requirements analysis, part technical architecture, and part project management, specifically estimation, and I have to say, Amen, Bruce!

Projects in which the business and IT folks fundamentally get along and respect each other can survive the inevitable surprises–requirements no one thought of, technologies that don’t play well with others, and estimates that were, well, estimates.

People can’t get along and respect one another unless they can communicate, and that takes a shared vocabulary. Bruce Maas recommends the IIBA vocabulary as being relatively simple to learn. It’s more of a meta-vocabulary, establishing common understanding of words like “requirement,” that then give you a language to talk about the actual problem. There are alternatives to the IIBA way, but I agree that it’s better than no shared vocabulary at all.

With positive regard for each other and a way to communicate, the suits and the geeks then have a fighting chance of agreeing on a definition of success, not just for themselves as suits and geeks but for the project they’re taking on, together.

In his role as CIO, Bruce Maas has proactively worked to build relationships (through social events) and shared vocabulary (by bringing in IIBA training), apart from any specific project. Not typical CIO stuff.

Give it a listen. The interview is less than eight minutes long.