the underlying problem

The world is more complex and unpredictable than it appears, and that complexity and unpredictability is increasing steadily. In complex domains, plan-driven and prescriptive approaches to performing and managing work are inherently fragile.


Increasing complexity driven by digital connectedness is causing our world to become increasingly unpredictable. The rise of Amazon, the Arab Spring, and the post-launch explosion of Pokémon Go are but a few examples of connectedness-catalyzed complexity and disruption.

This trend is accelerating, particularly in domains that are ripe for automation. Revolutionary tech is around every corner. With AI, 5G, and other technologies exploding everywhere, we have no idea what new threats and opportunities will emerge over the next 12 months.

“Agility” is a half-measure, and is relative. As sure as the sun will rise, your current level of agility will soon be the norm if it isn’t already. You will become fragile to the next wave, and it’s coming sooner than you think. 


Fragile things “break” when exposed to disorder. Most human-designed systems, from jet engines to project plans, are fragile because they are composed of parts with rigid or hidden dependencies. When one or more of those dependencies is unsatisfied, these systems lose their integrity and cease to function.

Many of us share the experience when a single failed part lands our car at in the shop, or when weather somewhere delays a connecting flight stranding us in a musty motel room overnight. Less obvious are the broken dependencies that wreck strategies, project plans, or other initiatives. When a changing requirement wrecks a project schedule, we can say the schedule is “fragile to changing requirements”.

We’ve come to expect that our machinations will work, but more often than we would like to admit, they are fragile.


Most of the systems and processes we develop are fragile because we think we are better at seeing interdependencies and predicting events than we really are. Modern human achievements, particularly in the physical sciences, have made us believe that we can analyze, model, and predict virtually anything. This shift toward scientism, in business at least, began with Frederick Winslow Taylor’s Scientific Management.

Plans, designs, and roadmaps are models that, when applied to complex domains, fail to provide much predictive value. Understanding and modeling simple and direct cause and effect doesn’t translate to domains where the number of subtle immeasurable cause and effect relationships are many, linked by invisible feedback loops and time delays.

Ironically, Agile conferences are packed with sessions about analyzing and synthesizing agile workflows, distanced from Taylorism in name only. Most agile frameworks merely replace one form of fragility with another. 

Fragile Agile

Agile software development as originally conceived significantly reduces fragility. However, modern adoption usually abandons mastery of agile values and principles in favor of “agile frameworks” that you can essentially buy and “roll out”. Good intentions and arrogance combined to produce agile frameworks, prescriptive and proscriptive systems of organizational structures, roles, practices, rituals to make you agile, subject to the terms and condition in the fine print.

As systems in their own right, frameworks are essentially complicated machines with some parts not included to be provided by you. Like any complicated machine, all the vital parts to be in place and in good condition for the machine to function at all. The problem is compounded when, even if you have all the pieces in place, the machine if forced into a culture for with it wasn’t designed.

A common example is the role of “Product Owner” in Scrum. An engaged, knowledgable, and empowered PO is a mandatory component in the Scrum machine. But what if your organization has a product management structure composed of functional business managers, each with a few BAs? Guess what? Even if you’ve done everything else right, until you figure out how to fill that gap, you don’t get to be agile. Sorry.

What we need is some kind of organic agility, without the frameworks.


Attempts have been made to escape prediction-based organizations and prescriptive process frameworks in favor of emergent organizations called “teal organizations.” These organizations represent the shift from a command and control to a distributed governance model that responds rapidly to local conditions, significantly increasing their resilience. Teal organizations are a step in the right direction but suffer from a few significant challenges.

First, they are implemented using a framework of rules which constitutes a complicated, and therefore fragile, “machine.” So like complex agile frameworks, most organizations will fail to successfully get all the pieces working.

Second, although they can change rapidly in response to stressors, there is nothing to ensure the changes that they make are good ones because they are subject to the same influences and biases that already cause people to make bad decisions. They can move quickly, but there is nothing inherent in their rules ensure that their movements are good.

Finally, the strong boundaries between structures and substructures block the formation of small-world networks present in every living organism.

What do we do about it?

Companies are closing their doors at an accelerating rate. Do we fall for another restatement of agile values and principles? Do we continue to chase one mechanistic management approach after the next? No—The answer lies in understanding and applying simple rules to counter the very nature of fragility itself.

Learn More
Share This