Wednesday, December 10, 2008

The Case for an Agile Fringe

In the background, the debate with the Agile 2009 Conference committee has been continuing since I raised my disappointment with the announced program. Olav Maassen and Eric Willeke have stepped forward and proposed an Agile Fringe stage that would encompass the ideas of the Breaking Acts and Questioning Agile stages from the 2008 event. However, the organizers are pushing back claiming both a lack of space and lack of need. Amhed Sidky is arguing that the existing program can accommodate the types of new ideas and agile skeptic proposals that the Fringe stage would actively solicit. I believe that the Agile Alliance and the Agile 2009 conference committee led by Johanna Rothman with program chair Ahmed Sidky need convincing of both the purpose, value and strength of support for an Agile Fringe stage within the main conference. So I am posting here an open letter to Johanna and Ahmed. If you agree with the sentiment in this letter I would like you to sign it by leaving a comment. Thanks.

Dear Johanna and Ahmed,

I greatly value the contribution you make to our community giving your time and energy to facilitate the gathering of our community of peers at Agile 2009. The Agile conference is the public face of the Agile Alliance to the wider community of software developers and technology industry professionals. So you have taken on positions of great responsibility. Your decisions about the conference send a public message to the wider professional community about the Agile Alliance and what it values. I strongly believe that the Agile Alliance values innovation, openness and objective debate about better ways of building software. I believe it is important that the Agile Alliance shows leadership and stands by those values and protects innovation, and shows openness through Devil’s advocacy by giving these topics specific space in the conference program. Two members of the Agile Alliance, Olav Maassen and Eric Willeke have proposed an Agile Fringe stage for the conference program to serve this purpose. I commend this idea to you and I hope that you both appreciate its value and show good leadership and judgment through its inclusion in the program.

Ahmed has argued in private email that such a stage is not required. He argues that the existing program can accommodate new ideas such as Real Option Theory or ideas which engender skepticism amongst the agile community such as Agile+CMMI. He also argues that the theme of this year’s program is to have domain specific stages e.g., a testing stage, a developer stage, a product management stage, and so forth, and that cross-cutting stage concepts are not part of the plan for Agile 2009. He argues that an Agile Fringe stage would be cross-cutting and therefore doesn’t fit neatly with the pattern. He continues that any submission for a Fringe stage could easily be accommodated in one of the domain specific stages. And further, that the review committees for these stages are open-minded, experienced professionals capable of judging submissions each on their own merits. That the open nature of the submission system will give innovative new ideas a fair shake and that there is no need for a specific stage to provide a home for emerging concepts or skeptical challenges to agile values, beliefs and practices. I would contend that this opinion is wrong-headed and not supported by history from previous conferences.

Others have argued that the open space area is all that is needed to accommodate innovative ideas and skeptical dissent. While there is truth to this I would argue that open space is not enough.

If we look at the emergence of kanban as a case study, kanban first started as a single open space at Agile 2007. At Agile 2008, there were 6 kanban presentation by other presenters, some of whom had been present at the open space session the year before. Almost all of those 6 kanban presentations appeared on the Breaking Acts stage. If Breaking Acts had not been present, it is likely that many of the kanban submissions would have been rejected. Rejected for entirely explainable human reasons. Rejected because they wouldn’t be interesting to a wide audience. Rejected because the reviewer didn’t know about or understand this new technique and found it confusing, threatening, or simply not interesting. As Chris Matts pointed out during the open review period for Agile 2008 in response to a reviewer comment, “we have 5 submissions about the specifics of prioritizing the Sprint backlog in Scrum, surely we have room for one more presentation on kanban…?” New things get rejected by reviewers looking for the familiar and the safe. Without a specific category for the new and challenging, it’s human nature to go with familiar safe choice.

A second valuable case study from Agile 2008 would be the treatment of submissions related to CMMI and organizational maturity. Many of the review comments were based on ignorance, fear and loathing and certainly not objective in nature. The treatment of the submissions and the submitters was often shameful and reflected poorly on us as a community and on the individuals making the submissions. I used my status in the community to dedicate a section of my main stage speech to the topic of Agile and CMMI. Afterwards, I received many positive comments and emails from regular attendees who were amazed that no one else was talking about it. This shows us that reviewers don’t often reflect the true concerns of the attendees. Reconciling agile methods with CMMI has become a hot topic outside Agile Alliance circles this year and the program for the Software Engineering Institute’s SEPG conference in March 2009 contains many presentations on Lean, Agile and CMMI including some from people who presented on the Agile 2008 Breaking Acts stage and some from those who were vitriolically rejected through the Agile 2008 submission system. Again, history would indicate that the existing program selection and review system for the Agile conference does not produce the right results.

While I have chosen to highlight Agile+CMMI and Real Option Theory as topics I believe would find a home on an Agile Fringe stage, I believe there must be several more out there, perhaps from people we’ve never yet heard of? How do we provide them with an audience and a hearing? If we allow the program to be driven by popular democracy we will always be trending towards the middle, the mediocre and the status quo. We will be reinforcing a narrow definition of agile and sending a message to our community that outsiders with new ideas are unwelcome.

I find Ahmed’s attitude to be academic and unrealistic. It relies on the process and removes focus from the people. His argument that all the submissions for Breaking Acts in 2008 could have been accommodated on other stages might be valid in theory but in practice they would never have been accepted. In the agile community we should know better than this. We value individuals and their interactions more than we value processes and tools. Ahmed is asking me to put my faith in the review process and the online review tool. I prefer to put my faith in human nature. If we do not provide a stage that explicitly solicits innovation, skepticism and ideas from outside the established and accepted notion of agile methods, we will have a conference devoid of significant discontinuous innovation. I would surmise that we have an echo chamber.

I truly hope that you will find a way to adjust the program to include the Agile Fringe stage.

Best regards,

David Technorati tag: Agile+2009, Agile+Chicago, Agile+Alliance, David+Anderson, Eric+Willeke, Olav+Maassen, Johanna+Rothman

Monday, December 08, 2008

Encouraging signs from Agile 2009

So while I posted my criticism of the Agile 2009 program a few days ago, I thought I’d take a moment to comment on the trends that I see as encouraging and favorable.

Most of all I am delighted to see a stage dedicated to tools! :-D It seems that the elephant in the room has been unveiled! The Agile Manifesto taught us to despise tools and processes. Despite this, from the beginning agile development has been about tools and processes. They were just different tools and different processes from those that preceded them [in the 1990’s.] So we went around saying tools and processes are bad, when what we meant was, your tools and processes are bad, our’s are good! So :-p :-p :-p to you! (traditional, waterfall, SDLC, academic, big process, big design folks.)

So I see a full stage dedicated to tools as a real coming out for the community. Meanwhile, the type of tools I see being developed in our community are quite astounding. I’m a particular fan of JBehave and the work from the BDD community. So I’m looking forward to what else shows up on the tools stage.

So what else looks interesting?...

I’m particularly glad to see a user experience stage. I was, after all, the UX guy on the Singapore Project when FDD was born. And most of my early publishing was agile user experience material. I wish the abstract for this program was a little more specific and gave better guidance for submissions but regardless, it’s good to see that user experience is getting a fair shake. It is so vital to the delivery of customer value. Technorati tag: Agile+2009, Agile+Chicago, Agile+Alliance, User+Experience, David+Anderson

Thursday, December 04, 2008

Clinton Keith on Why Kanban Makes More Sense for Games

Clinton Keith‘s name is associated with Scrum. His website even proclaims it boldly. But recently Clinton has been doing a lot more with kanban. He had a kanban/lean pull system presentation at Agile 2008 and now this Gamasutra article. It’s really well worth reading, though its heavy reference to manufacturing plants is not to my personal taste. Clinton reports that using kanban has shrunk lead time on lgame evel production activities by 56%.

In the example above, the team went from producing a level every 16 weeks to producing a zone every week. With seven zones per level, the ultimate improvement to level production was 56%.

This snippet provides some background ...

Asset creation is deterministic and sequential work that does not fit the Sprint iteration cycle very well. If we think about production as a factory assembly line, then the two to four week iteration cycle doesn’t make as much sense. Factories don’t empty the assembly line every four weeks and determine what to build next.

Assembly lines have things rolling off much more frequently and require incremental improvements instantaneously. The rate that completed assets roll off the line becomes the new heartbeat of the production team.

[...]

However, when we enter production, we can have a long chain of tasks that need to occur before we see some production assets in the game. Take, for example, the steps that need to occur for a single level to appear in the game.

Technorati tag: David+Anderson, Lean, Kanban, Agile, Games+Development, Clinton+Keith

Disappointed with Agile 2009 Program

Am I the only one who is disappointed with the lack of vision and leadership in the Agile 2009 program? The Breaking Acts stage is gone from next year’s event. There is no explicit place in the program for innovation that doesn’t already fit with the established notion of agile. After dedicating the largest share of my main stage speech in 2008 to this topic I am extremely disappointed that innovation and new agile method ideas don’t seem to be important to the organizing committee.

Equally, there is no explicit place for Lean in the program. Given the very evident shift in the community towards Lean principles and practices and the rapid growth of kanban, and the fact that the Gordon Pask Award was given to Kenji Hiranabe and Arlo Belshee explicitly for their contributions bringing Lean/Kanban ideas to the agile community, this is a truly surprising ommission.

But what surprises me more is the degree of duplication in the published program. What for example, is the difference between Agile Product Management and Customers & Business Value?

One of them asks us for “best practices for prioritization/ROI, getting customer input, roadmaps and releases, etc. Commercial products (for revenue) versus internal projects (for in-house customers)” and the other for “Defining and measuring the business value of projects, features, and processes. Using business value as a means of deciding which functionality to pursue. nventing, prioritizing, managing, and validating requirements. Working with customers and stakeholders Process/Mechanics”

And then there is Agile Adoption and Agile & Organizational Culture. What is the difference?

The latter asks us for “Agile is all about changing your organization. It’s not only about changing the way you think and work. Improving your organization’s agility in a sustainable way may also require changing its underlying values and principles. Change doesn’t come easy. An agile initiative doesn’t take place in a vacuum; it has to interface with the existing (organizational) culture. Both will influence each other as change takes place. This is a process of mutual adaptation, where one possible result is that the agile initiative can fail because the neither the organization nor the agile initiative is sufficiently adaptable.”

while the former suggests… “Knowing what Agile is quite different from knowing how to roll out an Agile process. The Agile Adoption stage will focus on lessons learned from rolling out an Agile process. [...] How can you assess your potential for becoming Agile before starting a migration? How can you minimize delivery risk when moving to an Agile process? How can you implement an Agile process that recognizes your unique business model, customer, and company culture? What does the Agile migration process look like in application? What are the key foundation steps for starting a move to Agile? What are the best practices for getting executives, managers, customers and team members to buy-in to moving to an Agile process? Can you mix Agile practices with your existing processes when migrating to Agile? Can you migrate too quickly? Can you migrate too slowly? How can you sell adopting Agile during the current financial crisis?”

Does one of these sound like a subset of the other?

And then there is a Telling Our Stories stage that asks for experience reports, but then so do almost all of the other stages.

I can’t help feeling that with some portfolio management of the program, the committee could have consolidated the existing stages and freed up 3 slots to use for something else. I personally would like to see the Breaking Acts stage back again. And I’d like to see a Lean stage. What else would you like to see at the Agile conference?

Update: Read Johanna Rothman’s reply.  Technorati tag: Agile+2009, Agile+Chicago, Agile+Alliance, Lean, Kanban, David+Anderson

Monday, December 01, 2008

Plans for Brazil and Argentina Changed

I’ve removed my post about Zen of Agile Management classes in Brazil and Argentina in 2009. The Argentina class is postponed until the 2nd quarter - probably June. The Brazil class is postponed and it’s unknown when it will take place. I’ll post a new note on this when I have more details - probably in early 2009. Meanwhile, if you are in Brazil or Argentina and are interest in taking my class on Agile Management or classes on Kanban or Agile+CMMI please get in touch.

Technorati tag: Agile, CMMI, Management, Leadership, Heptagon, Liveware