How I think about: Uncertainty

This is part one of a series of posts focused on a range of topics, concepts and ideas that, in my experience, significantly influence the level of overall success in any organisation. Some posts will be more about specific practices and approaches while others, like this one, will be more conceptual and aimed at creating a foundational understanding and a basic mental model of a specific idea. In this post, I hope to provide a perspective on the phenomenon of uncertainty and its different flavours and introduce some basic concepts and approaches which help to navigate uncertainty in business and in life.

The urge to ‘know’

As humans, we like to know. We like to feel we are in control. We want to know that we are in charge of our destiny — be it in our personal lives or in the context of the organisation we are part of. And we go to extraordinary lengths to reduce uncertainty. We analyse and compare scenarios, we devise intricate plans, and we identify and manage risks. We do all these things and more in a hope that we will reduce or ideally eliminate the uncertainty. And yet, despite all this effort, we continue to be surprised by what actually happens and how difficult it is to be truly ‘in control’ in reality.

The human desire to be certain, to be in control is natural and it stems from our evolutionary history. Being able to reduce uncertainty and predict the outcomes of our actions or decisions would be hugely valuable. In many areas of life, we pay experts to help us reduce uncertainty (think investment bankers, financial advisors, pollsters, political advisors etc.). Some people even pay charlatans to make them at least feel that they ‘know’ (think astrologers, psychics, fortune-tellers etc). And most of us spend money to reduce possible negative impacts of uncertainties in life (think insurance).

We crave certainty because the feeling of being uncertain is something deeply uncomfortable, distressing and altogether undesirable. It creates anxiety, makes decision-making harder, and forces us to keep changing our plans. Uncertainty can be so uncomfortable that we are prepared to fool ourselves into feeling confident or certain about something even if such feeling is not grounded in any meaningful data or evidence (think most religions). Moreover, an uncontrolled quest for certainty resulting in holding one’s opinions too strongly hinders progress, impedes innovation, perpetuates falsehoods, ignorance and injustice and shuts down conversations (think religious or cultural dogmas, Taylorism applied to knowledge work, superstitions, creationism etc.)

In a nutshell, feeling uncomfortable with uncertainty is very human. And while working effectively towards reducing uncertainty is logical and can be hugely advantageous in life and business, our discomfort with uncertainty can and does often lead to waste and creates real-world harm. Getting good at living with, managing and exploiting uncertainty, therefore, makes a lot of sense.

Where does uncertainty come from?

Uncertainty is all around us. The more you look, the more uncertainty you will find. In fact, even events that we normally tend to consider ‘certain’ (the sun will rise tomorrow) carry a degree of uncertainty. So where does all this uncertainty come from? At the most basic level, uncertainty is both a physical property of the universe (e.g., quantum mechanical events that can only be described in probabilistic terms) and of human society and the world we live in (e.g., moral uncertainty about what is the right thing to do in a given situation or the impossibility to practically anticipate all effects of all our actions). As we shall see later, while some uncertainties can be reduced, others can’t.

In a business context, uncertainties are inherent to running a business and, as such, they have many forms — political, economic, social, structural, financial, organisation, technical etc. As a result, businesses spend billions on dealing with uncertainty. Yet, in many cases, a lot of this investment is wasted and yields zero or even negative value.

Types of uncertainty

It’s important to recognise that there are several distinct types of uncertainty and dealing with each type of uncertainty requires a different approach and a different mindset. In addition, not all uncertainties can be meaningfully reduced or eliminated, some not even in principle.

The main types of uncertainty include:

• Epistemic uncertainty — being uncertain due to a lack of knowledge or understanding e.g., I am uncertain about the colour of the t-shirt my son is wearing today

• Aleatory uncertainty — being uncertain in principle e.g., when flipping an (ideal) coin, I am uncertain if I will get heads or tails.

• Ontological uncertainty — different parties in the same interactions have different conceptualisations about what kinds of entities inhabit their world, what kinds of interactions these entities have and how the entities and their interactions change as a result of these interactions.

• Semantic uncertainty — different participants in the same interactions giving different meanings to the same term, phrase and/or actions. The words and concepts that we are using are inadequate to describe what we are trying to explain.

In the case of epistemic uncertainty, if we gain more information (e.g., by doing research, taking more measurements, conducting tests etc) the uncertainty can be reduced.

Aleatory uncertainty, on the other hand, is irreducible in that there will always be variability in the underlying variables. These uncertainties are characterised by a probability distribution and need to be dealt with as such.

Ontological uncertainty is about different people (teams, organisations, social groups) having different mental models about the state of the world, and/or different perception of how cause and effect work in the given context.

Semantic uncertainty, similarly to Ontological uncertainty, results in communication and interaction issues, confusion, misunderstanding and conflict as people talk cross-purposes. The latter two types of uncertainty can be reduced e.g. by explicit definition of terms, their meaning and public ‘validation’ of these terms to create a shared language and taxonomy across a team or an organisation. One should also openly examine and discuss existing perceptions and mental models that different parts of the organisation have (e.g. using Causal Loop Diagrams and other techniques) and seek a common interpretation.

In organisations, we encounter all four types of uncertainty regularly, yet we rarely think more deeply about what type we are dealing with, and if and how can it be effectively reduced. Way too often we implicitly assume that all our uncertainties are epistemic in nature.

Finally, uncertainty is, of course, very closely related to the concept of complexity. Highly complex systems, organisations, processes, and interactions often tend to create or amplify existing uncertainty as relationships between cause and effect are unclear and emergent and most actors have only a limited understanding of the system as a whole. In such systems, most actions have a range of second, third and higher-order consequences some of which are highly uncertain. Meaningfully reducing such uncertainty often requires a more fundamental rethink of the organisational model, process, structure, technology, the flow of work or other core ‘properties’ of the system as a whole and reducing the overall complexity.

What can we do?

Organisations need to get much better at embracing uncertainty.

By ‘embracing’, I specifically mean taking steps to:

• recognise when a specific uncertainty exists

• understand the nature of the uncertainty

• clearly describe its nature and properties and carefully think about them

• assess the cost AND the value of reducing/eliminating the uncertainty

• where it is meaningful, effectively reduce the uncertainty (but only to the extent warranted by the cost/value analysis)

• establish a culture and technological, organisational, structural, and systemic capabilities which enable working effectively in the presence of uncertainty and respond to change

• create and foster organisational ’antifragility’ on multiple levels

“Not to be absolutely certain is, I think, one of the essential things in rationality”

— Bertrand Russell

The above should be read less as a checklist and more as a set of practices, qualities and organisational traits — a mindset about uncertainty. In this context, I would like to emphasise that reducing uncertainty comes at a cost. To reduce the level of uncertainty, we will need to invest time, effort, money, and other resources. This, of course, promises a reward in a form of greater confidence, better predictability, or more optimal and less risky decisions. However, it also comes at a cost which can be very significant. Way too often costs of reducing uncertainty remain hidden and are not explicitly talked about. This can result in an excessive (and ultimately futile) focus on uncertainty reduction (e.g., through more analysis, planning, detailed specifications etc) even though a focus on organisational, technical and process changes enabling effective uncertainty management and real organisational responsiveness/adaptability/agility would have been a much smarter and more productive strategy. Finding the balance is hard and, as is often the case, it all starts with the right conversation.

In practice, there are many ways, tools and techniques which enable us to work with uncertainty better. Context, as always, matters but, in general, when navigating uncertainty I find ideas like the Cynefin framework, Bayesian inference, Decision Theory (including basic concepts like Minimax, Maximin), Systems thinking, and Probabilistic forecasting very useful. When used in the right way with the right level of understanding of the uncertainty we are dealing with and the overall system or area we operate within, these approaches yield remarkable results and enable us to fundamentally change the conversation and the way we look at dealing with uncertainty.

Uncertainties, like taxes, are here to stay. Trying to completely eliminate them or ignore them is dangerous, can be very costly (and futile) and leads to bad outcomes in the long term. Uncertainties also create opportunities. Exploiting these opportunities requires teaching ourselves to live with, embrace and take advantage of uncertainties around us. It means getting comfortable with the discomfort, thinking more deeply about the uncertainties we need to work with and reframing the conversation we are having. Organisations that embrace uncertainty tend to reliably outperform those that don’t — both in the short and the long run. Good luck!

PS: Found an error? Or a typo? Did I get something wrong? Or do you have an idea you’d like to share? Please let me know!