It was naturally while writing edition II of Kanban with Laurent Morisseau that this topic of constraint, of good constraint, spontaneously came to me.

This covid pandemic was dramatic, but in all things there’s a silver lining—it allowed us to measure the value of constraint. Constrained by the injunction to stay at home, organizations developed their remote work capabilities. The constraint drastically reduced the variability of the answer to the question: can we, do we know how to, work remotely? The constraint increased predictability: we were going to know quickly. We weren’t going to remain in the fog.

I associate this with the revision of this book on Kanban, because constraint in Kanban is often used. In several ways, but the best known is the application of limits. We limit work in progress to ensure predictability. You have to push the envelope to really understand what’s happening. You are a team of six people, I constrain you to work on only one single item to deliver to me. This greatly improves my predictability. The constraint improves predictability. The item will be delivered in a fairly short time even if I lose overall performance. If I let the team work on ten items, I may improve my overall performance with better utilization of the effort and skills of the six people, but I lose the assurance of having at least one single finished item, and finished as quickly as possible. You see the idea. Let’s push the reasoning to the absurd: if I ask you to deliver nothing for this project, I have the certainty that the project is finished immediately, super predictability—absurd, I told you. Conversely, if I tell you: do whatever you want for this project and present it to me whenever you want, that is to say without any constraint, well I have no predictability.

Limit, frame, constraint—all of these are related. They are useful as long as they respect certain conditions.

The frame that describes what is expected of a group, of a team, in sociocracy or holacracy—its reason for being, its accountabilities, etc.—gives autonomy to the group, because it has a reference point, a meaning, rules, and it is autonomous, self-managing, as to how it responds to them. (How many times have I observed people confusing sociocracy or holacracy with absence of frame or constraints…).

Limits in Kanban for example: if we decide to limit the items we work on, it improves the predictability of our results (as mentioned above).

Constraints obligate. To think differently, to imagine unexpected things, to cross-pollinate concepts, to work together, etc.

The constraint of showing finished things every two weeks (or one or three, who cares), obligates us to know, to see, to share, and reduces ignorance.

The boundary between frame, limit, constraint is subtle, and doesn’t necessarily matter. Two things matter.

On one hand it improves my predictability. The sociocratic group defined by a frame will evolve within that frame (and won’t do something unpredictable, or at least will be oriented), limits or constraints reduce variability (it doesn’t go in all directions, I’m informed every two weeks, everyone focused on this item, etc.).

On the other hand it gives autonomy, it’s a way of governing that emancipates (under certain conditions we’ll discuss below). Because I’m not saying how to do it, but what is expected of a group subject to a frame, a limit, a constraint. Everything that is not the limit, the constraint, the frame, is possible. Normally it’s quite a vast space.

“Do what you want as long as in order to achieve this objective” is how I express it. Which has nothing to do with: “Do this to achieve this objective”. When we talk about engagement, or even principles for evolving in complex worlds, the first expression responds to these challenges. The second expression is management that will work in rather rare cases (violent and short crises for example) and will generally be counterproductive.

Thus the constraint, the frame, the limit possesses this second quality of freeing up a space, of engaging by giving autonomy and therefore a certain accountability, on the condition that it frees up a space. We free up this space by not saying how to do it, by not describing the expected gesture. We free up this space by not overloading it with constraints, limits, frames. If they are too numerous (constraints, limits), they suffocate and have ultimately become rules that impose gestures “Do this”, because there’s no space for anything else.

I think we have here the keys to good group dynamics, good project management.