The most interesting thing for many people will be that DAD is goals-based, and is far less prescriptive than even methods such as Scrum. For example, one goal is to manage change on your project. Empirically, we can observe that there are several strategies for doing so, which I described earlier, and they each have their advantages and disadvantages. So, where Scrum prescribes a product backlog approach, the DAD process framework gives you options and provides advice for helping you to decide which strategy is best suited for your situation.
Similarly, DAD provides a decision framework to address a host of other goals such as initial requirements modeling, initial planning, coordinating team activities (there are more options available than just daily Scrum meetings), integrating your work, and so on. So, instead of assuming that everyone is a process expert who can tailor up methods such as Scrum and XP or tailor down methods such as Unified Process to meet their needs, DAD takes the middle road and describes your options and the tradeoffs surrounding them. These options come from a variety of sources including Scrum, XP, Agile Modeling, Unified Process, Kanban, Outside-In Development, Agile Data, and many more.
Heather Shanholtzer: I understand you have been conducting an ongoing survey into agile enterprise practices. How is that going? Have you been surprised by the results?
Scott Ambler: I run a series of ongoing surveys, and new surveys are announced at www.ambysoft.com/surveys/. Survey results, including original source data, are also published there. The surveys are completely open and the results are being used by several graduate students in their thesis work.
Each time I run a survey there are always interesting and unexpected results. I have data on how agile teams are succeeding at scale, including all of the scaling factors I mentioned earlier. I have data showing that the majority of teams are, in fact, investing time up front doing initial requirements modeling and architecture modeling, regardless of what some people will claim. I also have data showing that agilists are more likely to model than traditionalists, something I wouldn’t have predicted, and that the deliverable documentation is at the same level of quality as that produced by traditional teams. I have data showing that some of the “cool” practices that everyone likes to talk about, such as test-driven development and continuous deployment, are nowhere near as common as you would gather from all the discussion around them, and many other interesting agile topics, too.
Heather Shanholtzer: How can readers learn more or participate in the survey?
Scott Ambler: I would love to have readers participate in the surveys, as it’s one of the best ways we have in this industry to cut through some of the rhetoric that’s flying around about agile approaches (both good and bad). The easiest way to do so is to follow me on Twitter, @scottwambler, and follow the link when I announce a new survey. Or you can keep an eye on my survey page.
Scott led the session Disciplined Agile Delivery: The Foundation for Scaling Agile at the Agile Development and Better Software conferences June 10-15, 2012. The book Disciplined Agile Delivery: A Practitioner’s Guide to Agile Software Delivery in the Enterprise by Scott W. Ambler and Mark Lines is now available. Visit www.ambysoft.com/books/dad.html for details.