David Anderson (@agilemanager) est l’un des instigateurs du mouvement Kanban pour gérer les développements informatiques. Nous aimons toujours demander aux figures emblématiques de notre écosystème de dispenser des formations, et pour ne pas déroger à cette ambition, David était de passage fin Octobre dans nos locaux pour donner une formation sur Kanban. Nous avons profité de l’occasion pour faire une courte interview video avec David et lui demander de nous donner sa vision sur le développement de ce mouvement prometteur. On y parle de proximité avec le mouvement agile, de gestion du changement, et si Kanban va ou non franchir le fossé dans sa courbe d’adoption. A visionner si vous n’êtes pas convaincu que cette approche va prendre une place prépondérante dans les années à venir au sein de nos DSI.
Interview de David Anderson sur Kanban par Gilles Mantel
Xebia: David, its a pleasure to welcome you at Xebia. I would like to ask you 2 main questions. The first question relates to the challenge of Kanban. I have read a lot about the challenges posed by Kanban and the extent to which it is agile or not, whether it can succeed. Could you give us your opinion on those perceived challenges?
David Anderson: I started the Kanban idea because I saw that certain organizations were struggling to adapt agile methods or they had rejected them completely. So my motivation was to find a way to help those organizations become more agile with a small « a », how to improve their business agility, to improve their economic results without necessarily forcing a recipe or a textbook upon them. Is Kanban agile or not? My perception of that question is that it has become very tribal in the way people ask it. They define agile based on practices performed rather that a certain set of values. So if you asked whether Kanban is a software development or an agile project management method, the answer would be no. What we are really doing with Kanban is helping people find a way to change. I would describe it as an agile change management method. It that sense there is nothing to compare it with. There is no agile change management method and there is no set of agile change management practices from the agile tribe. There are, however, some core values that the people who started the agile movement understood. One of those was that in the agile world, we make progress with imperfect information. The idea being that in the old world we tried to make the requirements and the analysis really perfect because getting changes in late would have been very costly. So they would basically make progress with imperfect information and refactor later on. So does Kanban allow you to make progress with imperfect information and refactor later? Yes it does.
The other core idea in agile is if I trust that you are a capable software developer and capable of doing a good job, then I should behave in a way to show trust in you and empower you to self organize and get on with doing what you do best. I shouldn’t waste a lot of time asking you for plans and commitments either and auditing you later. You should be empowered to do your own thing and self organize appropriately. So the question is whether Kanban allows for that high level of trust? Again, yes it does in fact.
The third idea is that knowledge work is in fact perishable and that storing large quantities of it in documents is not a good idea. It’s perishable and it’s going bad because of lot of the knowledge is tacit and is in the relationship between individuals. Therefore we want to shorten the time it takes to complete a task and minimizing the quantity of work in progress. That way we turn raw ideas into working software as quickly as possible. It’s quite clear that Kanban really helps with that. It explicitly limits the work in progress. It focuses on delivery times and it directly addresses the issue of knowledge work being perishable. So I think that Kanban is very aligned with the values that the original agilists had in mind all those years ago. Is it an agile method? Not as such. It is an agile change management process.
Xebia: so there are misunderstandings as to what Kanban actually is?
David Anderson : That’s right, there are misunderstandings. That’s understandable also because of the way people’s brains work. We like to put things in pigeon holes and when Kanban came along, people immediately jumped on the idea that it’s a new agile method. They started making comparisons which really were unfair and unreasonable but completely understandable because that’s the way our brains are wired up.
Xebia: Up until today you have had successes with Kanban. What future do you see for Kanban? Will it cross the chasm?
David Anderson: I don’t really know what the future holds. We’ll find out whether it crosses the chasm. When I look back at 5 years ago, there was one company doing Kanban and only a few people had even heard of it. 4 years ago there were probably 3 or 4 companies doing Kanban. There were still no book. There was only a yahoo group with 50 or 60 members. There were a few blog posts. 3 years ago, there had still been no conference. 2 years ago, there had been a conference in Miami, a conference in London with about 250 attendees altogether. We had a growing community and implementations on 5 continents. Now, in 2011, we have 4 conferences every year. This year in total, we’ll attract close to a 1000 people. We have a number of case studies from some of the world’s biggest companies with very successful results and enormous philosophy improvements. There is a well documented case study from the BBC for example with a huge velocity improvement with one team. Those things are very significant and there is momentum. But has it crossed a chasm? No definitely not. We have a few books, we have some case studies, we have lots of people coming to conferences but most people in the IT world have still never heard of it. Will it cross the chasm? I think that depends on whether we can show that Kanban solves real problems and helps people. It will cross the chasm if people are motivated by the problems it addresses and whether they believe that it fixes those problems. I’m determined to keep the community focused on pragmatic real examples and showing guidance that people can use and reporting real world examples. I believe if we keep doing that, we will see progress. Ask me that question in 3 or 4 years.
Xebia: Thank you for your insightful comments.