When do you not want an agile team?

For the past few years I’ve been an agile enthusiast, and I still am. But when we talk of scaling agile to the organization, it is easy to forget that agile practices work well under certain conditions. Many teams in an organization will not be able to create those conditions, and standard versions of agile rituals will not work as well for them as other “project management” approaches. This is old news, going back to the 1994 classic “The Wisdom of Teams” at least, but I don’t see it shared in most descriptions of “how to scale agile”.

In “The Wisdom of Teams”, the authors describe a productivity curve — the Y axis is productivity, and the X axis is how close a team is to being “real” or “high-performing”. The most interesting thing about the curve is that there is a substantial dip in productivity just at the start. To the left of the dip, a team is not really a team — it is a working group, with a clearly defined leader coordinating the work. The dip represents a period of self-organizing (often described with the classic Forming, Storming, Norming, Performing cycle). On the other side of the dip, we have teams with progressively more of the qualities we would consider to be “agile” today, enjoying ever-greater productivity.

The key is, if your team cannot work closely together in the same physical or virtual space, with constant easy informal communication, with every member taking responsibility for the work of the whole, then your team cannot make effective use of agile practices. Formal delegation and differentiated roles are necessary if the team’s ad hoc coordination takes more than a few seconds to arrange. Literally three seconds! I think the process of making a phone/VOIP call is long enough that it requires a context switch, and that means that the work stops. When I say “the same virtual space”, I mean working with the IM chat window open or conference call running, for hours at a time.

There are hacks available — for example, arranging work sessions where everyone is online at the same time, or working from the same office (or cafe). But unless this is the only way your team works, you will still need to divide up the work, which implies more structure than is usually part of “best practice” agile.

The approach used in sociocratic organizations represents a good example of how you can scale agile with a mix of team / working group types. Each team (in Sociocracy, a “circle”) meets at least once a month to do the team learning and decision-making that makes agile or lean teams into “guaranteed learning organizations”. At that time, they have the characteristics of a “real team”, setting their methodology together and taking ownership of their outputs as a team. Between these “circle meetings”, however, some sociocratic circles use agile ritual, and others use classic approaches such as PRINCE2, with a high degree of coordination and direction from an “Operational Leader”.

In sociocratic organizations, a double-link connects teams up and down the functional hierarchy. It provides guidance from the general teams to the more specific and provides feedback from the specific to the general. A lead link (the “Operational Leader” just mentioned) provides a downward connection from the more-general circles to the more-concrete ones; a measuring link (a “Representative” elected by the more-concrete circle, who participates in the monthly “circle meetings” of both lower and upper circles) provides an upward connection. Thus, each circle has a well-defined structural connection to the circles above it (and below it). With the structural feedback loop built into the organization by these two roles, each circle is tightly coupled and dynamically steerable — whether they use agile processes internally, or more elaborate formal structures.

Does my description of “when you can’t use agile” match your experience of working in various teams? Are you familiar with other ways of scaling agile which don’t require every team to use the same methodology? Let me know!

Moving Together: Teaching the Principles of Sociocracy through Physical Play

I enjoyed leading an experiential session for the IMTA (International Movement Teachers’ Association) in early June — at the moment I’m calling it “Moving Forward Together — Experiencing the Principles of Sociocracy through Physical Play”. I’d like to share some brief rough notes with you on how that process worked; please feel free to experiment with this program and let me know what you discover. Note that if you are not an experienced Sociocracy practitioner, or you are not experienced in leading groups in movement exercises, you may not have the necessary background to make full use of this material. Please feel free to contact me with questions. [Also, I make some generalizations in my description — please forgive me if I am inaccurate in describing you!]

My plan was to work my way through each of the core principles of Sociocracy — Consent, Circles, Double-Linking, and Election by Consent — in an experiential way, with a minimum of explanation.

I started with explaining my intention with the experiential workshop, and checking whether people were willing to explore with me. Nobody was unwilling; this was our first consent decision, which I noted and celebrated. I then shared a very brief explanation of the concept of Consent in Sociocracy — the willingness to move forward together in an experiment, on the basis that there is no visible danger that will prevent us from accomplishing a shared Aim.

Without taking more than one or two questions, I asked people to get into groups of three. Each group chose a Goal, a Mover, and a Sensor. The Goal would make an animal noise, and slowly move around the room. (I asked each group to consent to the animal noise selected by the “Goal” — this was intended to symbolize the standard practice of consenting to a proposed Aim). The Mover would close her eyes, and move toward the Goal until she reached the Goal. The Sensor would stay with the Mover (perhaps touching) and say “safe” or “danger” as they went. Participants seemed to understand the instructions, and began play quickly.

As they moved, I informed them that once they caught the Goal, they would all switch roles.

After several rounds of this, I gathered people for Q and A in the large group. Predictably, people had made lots of useful discoveries. I did my best to answer questions, share my experience, and ask more questions.

Then I asked the groups of three to join together into groups of six. I gave each group of 6 a pillow. Now the pillow represented an Aim, a specific piece of value to deliver to the Goal. Each group of six was to choose a Goal by Consent. Then the Goals moved off into the room, and each team (circle) linked hands, closed their eyes, and followed. I had suggested to them they would need to call out ‘safe’ or ‘danger’ as they went, and this was essential to avoiding the various pillars and tables in the large space.

After several rounds of exploring in groups of six with some variation, I gathered people in the large group for more reflection. We explored what it was like delivering the pillow together, and how people felt when they all had their hands on the pillow, versus how they felt when only some of them did.

In any case, after a round of reflection, I asked them to double up their groups of six into groups of twelve. But before doubling up, I asked them to each elect a representative who would provide a link to the other group. I think this may have been the least useful element in the physical exploration, because I am not sure they really experienced the value of the double-link. In the future, I might try having the groups connected in different hand-holding patterns, to symbolize relationship patterns. This would clearly demonstrate that a two-person link is more robust and easier for coordination than a one-person link. In the exercise as we did it, I believe the double-link principle was mostly lost.

In any case, as they did multiple iterations of the game with the larger groups, I invited them to consider together between each round whether they wanted to make any changes to the game to make it more fun, a better learning experience, or just to help them feel more alive. This produced a variety of responses — some groups got more effective at delivering the Aim but found they were having less fun, which provoked the question for them of what their true Aim was. Several group noticed that if they consented to a new experiment quickly, they were able to spend more time playing and less time talking, even if the experiment wasn’t the one they most personally favored.

After several rounds of play in 12-person groups, I called everyone in for a final large circle discussion. I invited people to share what they learned, and reflected alongside them on our learning. The responses I heard from the group were strongly positive — they had enjoyed the playful exploration, and they felt enthusiastic about using such a dynamic system for steering their organization. The applause was warm — although not as warm as the applause following our next session, when we used the sociocratic method for electing new board members and making some policy decisions in their General Assembly. They enjoyed the experiential play, but they appreciated it even more when they saw that they could actually work this way.