Friday, July 27, 2012

Self-organization

We have this idea in Agile, that the Team should self-organize.  This is an important idea. And also an important occurrence (eg, the reality precedes the idea).

In Agile, self-organization is compared to command-and-control.

We think self-org is an important thing to study, both in general and in your Team.

Why

Well, one, because it is just right.  People are free, and self-organization is saying that the Team is allowed to be free.

I guess this needs to be explained a bit.  Some will say: well, the company has bought their time as employees, so the company gets to define what the Team does.  Well, let is concede this at a high level; let us say that the company may define the vision or the goal or the general product.  Perhaps even the user stories.  The Team members are not slaves, but we can assume that, as employees, they agree to do the company's work for pay.  A contract.

But, devising the work, figuring out how they will get to the goal, they should have the freedom to do.

The second main answer is: self-organizing humans tend to do better work than human 'slaves' (humans directed by one or a few command-and-control people).

There is lots of evidence of this. The first book I recommend is The Wisdom of Teams by Katzenbach and Smith.  This is a rather old book.  But the basic evidence is that a self-organizing team out-performs, almost always, an individual.  And to such a degree that the extra cost is well worth it.  But, you must given them some degree of 'freedom'.

There are also a lot of more recent evidence.

Some topic you may wish to research:
Self-organization
Complex Adaptive Systems
Maneuver Warfare
Free Enterprise
Knowledge Creation

Note: Some people say that a free enterprise system is mainly characterized by (mostly) private ownership of capital (means of production). And attribute its successes or failures to that. Others focus on the freedom of individuals to make their own decisions (much like the "wisdom of crowds" idea) and on the view of the national economy as a complex adaptive system, trying to accomplish high-level economic success via many individual agents (people or companies) making their own decisions.  In our view, successful free enterprise countries exhibit many of the successful characteristics related to self-organization.

Command-and-Control Managers

Lots of managers have been taught, explicitly or implicitly, that the 'workers' are dumb, and the manager must tell the worker how to work.  In our view, especially for virtually 100% of knowledge workers (our domain), this is very very incorrect teaching.  But it is nonetheless, what they have been taught.

Now, some of them also understand freedom to some degree, and understand that they want the people in their group to 'think for themselves'. But, we can say with relative assurance, most companies have a lot of people who are relatively command-and-control in their style.

And, in pressure situations (our common situation), they want to use too much command-and-control.

Again, this is not true of all managers, but of many. It depends, in part, on the company's culture.

It is hard to convince these people to be patient for self-organization.

Teams that won't self-organize

This is seen in real teams.

There seem to be several root causes. One, the team has been beat down by command-and-control managers so much, that they have 'forgotten' how to self-organize.  Another: that they Team does not believe it when managers say 'self-organize'.  Another: The Team is fearful that by self-organizing they will be 'held accountable' with punishment a likely outcome.  To avoid pain, they refuse to self-organize.

Whatever the reason, two things can be said.

Many Team (perhaps with Agile coaches) do eventually break out from being 'stuck' (not self-organizing).

And some Teams may take a very long time or perhaps never do it.

The key advice is this: If the Team will not self-organize, or seems to want to self-organize to mediocrity, then a good manager should intervene.  Temporarily.

Perhaps prior key advice: Be patient. Often Teams will self-organize in two or three sprints.  Keep talking about it, and ask them if they need help.

Third advice: In the view of some (including me), some Teams lack some key skills to self-organize. For example, a new junior Team may not know really how to break down work into tasks for the sprint planning meeting. Sometimes 'holding their hand' as they mature is a very successful approach.  And 3 or 4 sprint later, they can be very good at self-organizing their own Sprint Backlog.

Learning to decide as a Team

I think each Team learns how to decide.

The first, most useful thing, is that everyone in the Team gets to offer input on a decision.

Next, the Team needs to accept that no decision is ever perfect.  Thus, decisions in general must be made, perhaps not always quickly, but promptly.

The Team needs to understand the impact of decisions on the morale of the Team and on the success of the Team.

In our view, most teams go through a forming and storming period of decision-making.  And then later get better.

Good self-organization... and better

Lots of Teams seems to self-organize well.

What is also common is for most Teams to plateau.

In part, we may say that a root cause is that most Teams want to reach a stasis... a place where things are balanced and where change slows down.

But have they reached the height of improvement?  We think not.  Some Teams have reached much higher heights.  And some Teams keep on improving.

There seem to be two main factors (or sets of factors):
* magic -- or, more accurately, a bunch of things that are hard to describe.
* a bunch of factors that people talk about, and some teams study and work on

Some of these last factors seem to be quite 'soft'.  Love, listening, creativity, heart seem to be among them.

Lastly

If you have never been on a good team, it is hard to understand what self-organizing is all about.

Within the Team, it might be rather rough and ready. But a lot depends on the specific individuals in the Team.

No comments: