Roots of Scrum: Takeuchi and Self-Organizing Teams

The first Scrum team was created directly from a paper which is required reading for any Scrum practitioner. It explains how to set up self-organizing teams and clearly outlines management's role in the process.

Takeuchi and Nonaka. The New New Product Development Game. Harvard Business Review, 1986

A longer paper from a book which is out of print goes into more depth. Some may find this easier reading.

Ken-ichi Imai, Ikujiro Nonaka, and Hirotaka Takeuchi. Managing the New Product Development Process: How Japanese Companies Learn and Unlearn.

Also see Takeuchi and Nonaka's books. Their latest books do in-depth analyses of Toyota.

Takeuchi and Nonaka clearly explain the fundamental concepts so often missed by management and teams new to Scrum. The discussion of self-organizing teams is a good example:

A new product development team, consisting of members with diverse backgrounds and temperaments is hand picked by top management and is given a free hand to create something new. Given unconditional backing from the top, the team begins to operate like a corporate entrepreneur and engage in strategic initiatives that go beyond the current corporate domain. Members of this team often risk their reputation and sometimes their career to carry out their role as change agents for the organization at large.

Within the context of evolutionary theory, such a group is said to possess a self-reproductive capability. Several evolutionary theorists use the word "self-organization" to refer to a group capable of creating its own dynamic orderliness. A recent study by Burgelman found that a new venture group within a diversified firm in the United States takes on a self-organizing character. Another study by Nonaka has shown that Japanese companies with a self-organizing characteristic tend to have higher performance records than others.

The creation and, more importantly, the propagation of this kind of self-organizing product development team within Japanese companies represents a rare opportunity for the organization at large to break away from the built-in rigidity and hierarchy of day-to-day operations. It is quite difficult for a highly structured and seniority-based organization to mobilize itself for change, especially under noncrisis conditions. The effort collapses somewhere in the hierarchy. A new product development team is better suited to serve as a mote for corporate change because of its visibility ("we've been hand picked"), its legitimate power ("we have the unconditional support from the top to create something new"), and its sense of mission ("we're working to solve a crisis situtation").

An alternative view of a parallel approach is outlined in:

Chet Richards. Certain to Win. XLibris Corp, 2004.

Richards outlines the strategy of fighter pilot John Boyd. He uses German concepts to illustrate key ideas.

* Enheit -mutual trust, unity, and cohesion
* Fingerspitzengefuhl - intuitive feel, especially for complex and potentially chaotic situations
* Auftragstaktik - mission, generally considered a contract between superior and subordinate
* Schwerpunkt - Any concept that provides focus and direction to the operation

He takes these concepts down to second order interactions. Trust builds by training in the field and by mission driven challenges that a self-organization team accepts or does not accept. It they accept, they can be counted on to do it or die in whatever way they see fit without intervention by the superior. This level of trust depends on the team trusting the rightness of their leader. If superiors do anything to disrupt the team, trust breaks immediately.

With trust based on real unity and cohesion, Boyd's Observe, Orient, Decide, Act feedback loop goes into an implicit state where there is Observe, Orient, and Decisions becomes implicit. The team goes into motion before the leader can give a command. Like in martial arts the Sensei is moving before he even sees the motion of the attacher using a sixtth sense. This is the kind of trust a Dream Team has. You know it when you see it and how come so few software teams have that level of trust?