Cadence and granularity represent concepts where we get some strong links between Cynefin and Kanban. In this context I am using cadence in the sports sense of the world, the revolutions per minute if you cycle for example. Here it relates to bio-mechanical efficiency and maintaining a constant cadence using the gears is one of the mantras of long distance cycle. Granularity references the size of the things you are trying to manage – tasks if its Kanban, tasks, interventions, experiments etc. if its Cynefin. My three basic rules for designing complex systems interventions are (i) optimise the granularity (ii) distribute cognition and (iii) disintermediation of decision makers; granularity comes first. Within the complex domain of Cynefin with its focus on safe-to-fail interventions reduces the granularity from fail-safe design and thus reduces the danger of unintended consequences (more on that tomorrow) and thus increases resilience. Both Kanban and Cynefin recognise the cognitive aspects of granularity, too finely grained tasks and it is impossible to gain a visual sense of backlog, to coarsely grained and completion becomes exponentially impossible.
There is another aspect to this question, which also relates to dynamics within Cynefin. As I keep reminding people Cynefin has two typologies, one of type or domain and the other, patterns of dynamic movement between domains. The most stable dynamic is the move from complex to complicated, and back again. Experiments result is consistent results (complex to complicated), tests for continued consistency trigger a return to experiments if they fail. Different areas of activity will have a different cadence, a pattern of frequency in this transition. Shifting to the complicated domain is desirable as it makes scale easier and control more consistent; but only if there are sufficient constraints of the right nature in place to sustain that consistency.
This is where we get back to granularity. By reducing the granularity, you increase the likelihood of being a able to move activities into the complicated domain. Coarsely grained actions, more or less inevitably have complex elements and thus cannot fully make the transition. I want to emphasise that the transition is important if, and only if, it can be achieved. There is nothing innately good about being complex, let along chaotic despite what some enthusiasts would have you believe. How we visualise it however is another affair – that is where I think Kanban in its current practice (although not its concepts) cannot be a true complex system technique. For that we need different visualisations and possibly different mathematics. In a complex world tasks may not have knowable boundary conditions for example. So the idea of containers and connections in my growing theory of constraints is also important. Dark container, dark connections and there is no way a Kanban board has any utility; rigid container, variable connections maybe.
There is also an interesting aspect of symbiotic market life cycle management (more on that in a few days time as a development of my work on Apex Predator theory) where coupling ‘grains’ can allow for higher levels of stability at lower risk; in effect linking over the domain boundary to limit variation in the complex. This does need more development however so feel free to ignore this final paragraph for the moment.