Did you catch that? A PMI board member, talking about Agile?!
Now, he gave a similar talk to the New York City Chapter earlier this month. But letâ€™s be honest here, it doesnâ€™t get any more non-Agile than Washington DC, the home of large federal waterfall projects. So you can imagine the sense of anticipation and showdown that was leading into the evening event. Well, he certainly delivered, showing off a wry sense of humor and a brief moonwalk demo.
He introduced with a challenge to the group to consider that reflective criticism is not bad. PMI has an obligation to consider industry trends in PM best practices. The current PMBOK Guide is over 400 pages long, and is at risk for growing even larger, without adding any value. However, to defray the risk of his own PMI defrocking, he jokingly quipped that he would also be critiquing the CMM and ISO 9000 standards in light of the Agile trend.
Dr. Frame continued with the observation that Agile techniques are not completely a new phenomenon. Indeed, he “felt a little like Rip Van Winkle” when he got pulled into Agile Project Management a little over a year ago. Pioneers like Fred Brooks, Barry Boehm, et al. were advocating non-linear project management in the 60â€™s and 70â€™s.
He also described his own analysis of 100 RAD projects, where he found that not a single RAD project was customer-rejected. By engaging customer feedback during the development of a prototype, you can be sure that what you deliver is what they wanted. Traditional RAD has a prototyping phase followed by a waterfall implementation phase. So, you get dynamic “agile” feedback during the highest-risk phase of requirements gathering. In fact, the prototype can be so compelling that, in some cases, it gets deployed before official development has begun.
He highlighted the common sense practice of â€œOpportunistic Schedulingâ€, and Agileâ€™s heavy use of it. If a tester is available sooner than planned, wouldn’t you go ahead and use him, rather than wait for the planned start date? If a piece of hardware arrived a few days early, certainly you would have the team take advantage of it if they could. The plan is a guide, not a law. However, he asserted that the spirit of the PMBOK was to discourage such things, even if the letter of the guide actually permitted it.
In his overview, he grouped modern Project Management methods into three broad approaches: Waterfall, Iterative, and Agile. He observed that sometimes, there can be overlaps and combinations. For example, iterative methods can resemble successive mini-waterfall phases, or aggressive waterfall projects delivering frequently can look agile.
Towards the end, he walked through the 5 levels of CMM and contrasted their definitions with the philosophies of Agile. For example, having repeatable and defined processes is proscriptive and predictive in nature, which goes against the Agile tenants of adapting to your context. Interestingly, many Agilists disagree with this perspective, and assert that methods such as Scrum can be CMM compliant.
So at last he arrived at the final slide, declaring that the best approach was â€¦.(drumroll)â€¦ all of them. Each approach brings different strengths to different environments. Management practices should be situationally specific. Indeed, cultural baggage and personality are legitimate factors. Some project members simply canâ€™t work in a regimented environment and some will shrivel up in a highly dynamic environment.
Afterwards, I asked him his thoughts on those that believe you can reconcile the Agile techniques with the PMBOK (I would be one of those people). He disagreed, and only half-jokingly quipped that anyone trying to characterize PMBOK standards as Agile should be branded as a heretic. Wowâ€¦not exactly the response I would have expected from the guy who led the development of the original PMP exam. He agreed that the PMBOK motto â€œprogressively elaborationâ€ is a good tenant, but that the spirit of the guide is becoming more and more of a legislative tome, than an effective guide in PM best practices.
In the end, I am excited that Agile has gotten the devoted attention of a renowned expert and PMI insider like David Frame. However, I would contend, as I have in the past, that we do not have two mutually exclusive philosophies or communities. Indeed, the Agile method of Scrum has been shown to be fully CMM-3 and partially CMM-4 (see slide 99 of the Scrum intro course). PMI Global Congress speaker Mike Griffiths has shown that Agile focuses on project execution and control, where as the PMBOK devotes most of its process steps to planning.
All in all a very interesting event, and one that will motivate me to continue the conversation.