Learning to Anticipate Change

Learning to anticipate change is a challenge in complex environments. Business intelligence and predictive analytics can help make sense of the signals.

We have all heard, "A smooth sea never made a skilful sailor". Well, for an analyst, a complex sea of data makes the skilful means of sensing evermore necessary. An analyst also needs a crew to help anticipate and navigate the changes that may arrive. These changes arise in three types of systems: simple, complicated, and complex.

Simple systems trigger linear changes which are steady, wave-like, and easy to learn. Complicated systems may cause a vast array of effects, but if understood and reliably modelled, offer predictable outcomes. Complex systems are notoriously unpredictable, so give rise to non-linear changes that lack a clear basis in cause or effect. Liviu blogged:

…in complex environments cause and effect relationships do not repeat and a categorization mindset where data is fit to preconceived notions about reality (i.e. models, frameworks, etc.) is ineffective.

Dr. Will Allen, from Learning for Sustainability, notes that complex adaptive systems (CAS), such as human beings, stock markets, and ecosystems, "defy attempts to be created in an engineering effort"; furthermore, there are neither "simple [nor] complicated approaches to evidence" that will clarify their outcomes. Instead, Allen suggests that "well-thought-out" interventions be employed to gain insight into their operations. He shares a similar view from Irene Ng’s post on Complicated vs Complex Outcomes:

… we have spent the last 100 years doing complicated rather well. “We can pat our backs on putting the man on the moon, doing brain surgeries etc. We are now moving to a world where complex outcomes matter and this is a new capability. This capability uses different words. We can determine complicated outcomes. We can only enable complex outcomes. We can specify complicated systems. We can only intervene in complex systems.

What is an analyst to make of all of this if existing frameworks fail? And how can this “new capability” be gained?

Predictive Learning

In Syrett and Devine’s book on Managing Uncertainty, Predictive Learning is defined as "the capability to sense, probe and analyse previously hidden patterns and trends in order to anticipate sudden or disruptive change." In this book Adrian Rawcliffe from GlaxoSmithKline recommends that organisations seeking to develop predictive learning require the systematic process of "baked in" business intelligence, as well as the informal process of learning from an uncertain environment. Rawcliffe stresses that systems can’t be set up to gather the “unknown” stuff, therefore “external sensing” is required at all levels of an organisation.

Sensing what? Signals.

A friend and complexity expert, Dibyendu De comments, "the critical question to answer is, 'What is a signal anyway?'’ Syrett and Devine cite a definition of a signal by the +Institute for the Future (IFTF), which simply put, is a phenomena such as "an event, a local trend, or an organisation" that emerges from one scale and locale to potentially disrupt a larger one. The IFTF adds that "signals are useful for people who are trying to anticipate a highly uncertain future."

An organisation can wilfully learn from the signals that shape the future. Today, tools such as business intelligence and predictive analytics can certainly help make sense of simple and complicated situations. But a complex sea of data calls an analyst and crew to sense, probe, and analyse by applying predictive learning, the underlying capability to #AnticipateChange.

This capability will be explored in future posts, but for now a quote until we meet again.

"We must free ourselves of the hope that the sea will ever rest. We must learn to sail in high winds." ~ Aristotle Onassis

Image by the_tahoe_guy