Tuesday, August 08, 2006

Thoughts from Agile 2006

So I’ve had a couple of weeks to digest my thoughts from Agile 2006. I thought I’d share. I have a couple of big takeaways and a few other observations…

Scaling Agile. The BIG issue for this year is scaling agile across a whole organization. I see this as having three parts - program or multi-project management and the rollup of schedules and resource plans to a Director or VP level; architecture and enterprise level modeling of a domain and data center; and finally configuration management including build, integration, branch and merge strategies, and work-in-progress batching and related communication.

Agile Process Auditing. Another common theme on the cocktail party circuit was the story that goes something like this…

A project in our organization just failed. The trouble is it was allegedly an agile project and now our VP wants to kill all the agile projects.

After asking a few questions, it seemed that there was a need for an audit of the conformance to agile process so that the executive could be given all the facts.

The idea of auditing is anathema to agilists - it smells of big process and bureaucracy. Audits imply “low trust” and agile is all about “high trust.” So, is it reasonable to consider the idea of agile process auditing and quality assurance checkpoints?

Personally, I think it is. It is hard to scale trust across a whole organization and while it is maturing it is natural to give out trust in small amounts. We do this with children as they mature and grow - we simply don’t trust them to cross the road on their own until they are old enough to demonstrate that they can follow the rules and not get killed doing it. We constantly audit our children for trust compliance issues. So why not audit our infant agile teams to see whether they can walk without falling over?

And a few other asides…

On the first Sunday, I was approached by three different people who all wanted to tell me about how they were using Theory of Constraints to deliver great results in their organizations. They very much view me as the leader who introduced the ideas in TOC to software development and I am really excited that people are picking up the ideas and running with them and taking them to new heights.

The APLN Leadership Summit was a huge success. I’ve resigned from the board but have volunteered to be the blog evangelist for the APLN. There aren’t enough project management bloggers. I will be working to encourage more of them and hope to make AgileManagement.Net a hub for project management bloggers over the coming year.

After my session on FDD on Monday, I was mobbed with people asking about more FDD sessions at the conference - there weren’t any - no FDD paper has ever been accepted for any of the conferences over the years. Or failing that sessions on domain modeling, or architecture and agile and so forth. I see this as part of the pent up demand to solve the scale problem. FDD was designed to scale from the beginning. It was developed on a large enterprise, mission critical system. There is a lot the agile community can learn about scaling agile from reading the FDD literature. FDD has guidance on reporting up to the President level in the company - with the parking lot chart - assessing the ship readiness of the product - with the parking lot chart - scaling the architecture - with Coad Method modeling - and configuration management was built in to the method from the beginning with the build and promotion group system designed by Terry Gliedt.

And finally, I was hoping that we’d start to see the breadth of topics at the agile conference broaden into areas like agile UI design, agile database design and development, agile testing, agile analysis, agile architecture - but I’m really not seeing it. These topics come up in an occasional paper but are failing to gain traction and year-on-year momentum. What I am seeing is increasing depth and a focus on narrow esoteric areas of pair programming or TDD and so forth. My feeling is that if this trend continues the agile community will fail to deliver on the big opportunity - to fundamentally change software engineering and our profession for the better. So next year, I’d like to see the organizers encourage more breadth and bring in some new blood, some new topics and some greater heterogeneity to the event. Technorati tag: Agile, David+Anderson, Agile+2006+conference, APLN