The Amber Compass guides us to navigate in the complex world. It includes, from the centre outwards, a framework for understanding the nature of complexity, focus areas to pay particular attention to, and activities to engage in.
The purpose of the Compass is to build competence, instead of providing recipe-books. Behind the Compass, there are at the moment over sixty individual training units and tools, each with content and appropriate didactical methods, which allow us to customise an appropriate intervention to help your team or organisation become more effective in mastering complexity.
Our didactical backbone for exploring the framework and the principle on the compass is the Cynefin Playing Cards.
The Elements of the Amber Compass
Core: The Cynefin Framework
Dave Snowden’s Cynefin Framework is a basic sensemaking model to distinguish between various kinds of ordered, complex, and chaotic contexts. In the Compass, it serves as an anchor from which to enlarge our mental models about the nature of complex adaptive systems, and their fundamental difference to ordered contexts.
inner Circle: Focus areas
DISTRIBUTED INTELLIGENCE: A complex system can only be perceived partially. Different points of view produce varying and contradicting evidence. There is not one “right” description, but your own, single-source description is most probably insufficient. Following this principle, you will want to keep diverse sources open, and avoid excluding contradicting perceptions. When you take action, you will pay more heed to one side of the contradiction – but keep in mind that you act on a bet, not a certainty.
REMOVE FILTERS: Our perception of a complex system immediately filters out much information in order to create a coherent idea of the situation. Often, decisions are taken at a point where information goes through multiple layers, such as people, monitoring systems etc. Each layer applies a filter which seems coherent in relation to the information it perceives. But since nobody knows the nature of the filters previously applied, the multiple layering of filters will distort the information. You will therefore want to learn as much as you can about your own relevance filters, knowing that your understanding will never be complete. If possible, you want to keep decision-making and first-hand information close together.
ALIGNED AUTONOMY: Delegation of autonomy to the point of contact enables an organisation to deal with complexity, in the form of unexpected and exceptional cases. However, it is in the interest of the organisation that the multitude of individual interactions follow a common trend, so that as a whole, the organisation behaves with an adequate degree of focus and unity. You will therefore want to balance a high degree of autonomy with an effective common mindset or orientation.
ADJUST GRANULARITY: Perception and action in a complex system are partial, changing, unpredictable. If you want to cross a wood, you need to have a general, vague idea of the wood, but for your next steps you need concise knowledge of some trees around you. Or think of a swarm of bees instead of one eagle. Competence implies to integrate various granularities with all their imperfections, to switch between them appropriately, and to decide where you better invest your resources for the next phase of perception and action.
RESILIENCE: Change, surprises, accidents and failure are inevitable parts of acting within a complex system. Robustness is the ability to survive crisis unchanged. This rigid quality is very expensive to uphold in a complex environment. Resilience is the ability to survive crisis changed and most often more adequate.
outer circle: activities
NUDGE FOR EMERGENCE: A nudge is something small and non-coercive that can lead to a larger, desired behaviour in the system. In a complex environment, linear interventions in the sense of one cause – one effect are both unreliable and highly inefficient. Nudging tries to mobilise the potentials inherent in the system. Think of bringing your child into an interesting environment and then stepping out of the way.
EXPLORE-EXPLOIT: In most human-designed systems, productivity depends on the exploitation of order: standardisation, industrialisation, scalability. Imposing an island of order within a complex environment can make sense. At the same time, you need to constantly explore whether the patterns on which you rely start shifting, in order to know what to do next. Excessive focus on exploration is called eternal boiling and will not get you productive. Excessive focus on exploitation is called premature convergence and will eventually miss the moment when your assumptions are no longer correct. You need to do both, in adequate distribution of your resources.
MAKE SENSE: In a context of partial understanding, diverse points of view, and high autonomy, making sense of your environment becomes a constant, never-ending activity. If you are in an anthropo-complex environment, e.g. a consumer market, all the agents are constantly engaged in sensemaking, and you will want to learn what sense they are making of the system. If you are acting as an organisation, the internal sense-making needs to be done in common, with adequate degrees of variety and commonality, of centralisation and de-centralisation.
MANAGE CONSTRAINTS: Constraints limit what a system can do. They can be physical (walls, streets, doors), virtual (firewalls, budgets, laws), or mental (tradition, values, information). Depending on your position in the system, you can influence, or even impose constraints. You can make constraints narrow or loose (declare a solution, suggest a solution, brainstorm for a solution), or you can make them rigid, elastic, or permeable. Thus you influence system behaviour (e.g. speed, focus of attention, conformity pressure, commitment).
EXPERIMENT: You will not know what works unless you find out. Where analysis-ideation-implementation is too rigid, trial-and-error is better. You will want to test any coherent assumption as quickly and as cheaply as possible, and integrate the learning into your evolving practice.
MANAGE BY VECTORS: In ordered systems, you were able to manage by objectives: describe the result you want to see at the end of a period, and delegate the way to get there. In a complex system, you do not know now what result is best in the end. What you can do is describe a vector, based on the potential you observe in the present system: it gets hotter, so we want to go North – and that means avoiding South. You can observe alignment, and even speed of progress, while you delegate the concise management of how to overcome the next obstacle. Variants of vectors are heuristics and rules of thumb – something we all use in everyday life, while we rarely use them consciously in management.
FACILITATE RECOVERY: In the spirit of resilience, you want to decrease your system’s overall vulnerability by increasing its healing powers: contain damage, de-couple processes, increase redundancy, awareness, responsiveness, risk conscousness, risk monitoring, etc.Back to parent page