The illusion of control: Why IT leaders cannot rely on clear roles and responsibilities

As humans, we crave certainty. It creates predictability, a sense of safety and security in knowing how to succeed.

It’s no surprise that this instinct carries into the workplace. It is fair to expect employees to ask for clarity around roles, responsibilities and expectations, especially in a world where technology, markets and even jobs shift quickly.

While it may be human nature to seek certainty, as a technology leader, I’ve learned that clear roles are never the answer. We’re operating in times of unprecedented technological uncertainty and instead of trying to eliminate that uncertainty through clearer divisions of work, we must focus on better equipping people and organizations to handle it.

As leaders, the most valuable thing we can equip our teams with is resilience to withstand the discomfort of uncertainty and the empowerment to think creatively, adapt quickly and stay focused on the desired outcomes, regardless of how uncertain they may feel.

Understanding uncertainty

Uncertainty can be found in two dimensions: how knowable (what we know and don’t know) and how controllable (what we can or can’t do) our environment is.

When leaders set goals, designate roles and delegate responsibilities on the assumption that most things are known and that almost everything is in our control, we should be expected to make accurate predictions.

But reality is far messier and getting much less predictable. In business, you might not know what regulations or technological advances are just around the corner and the actions of competitors aren’t controllable.

A lot of the confusion we experience stems from efforts to satisfy both the desire for clarity and the need for structure while managing by the principle of focusing on what you can control and not wasting energy on the things you cannot.

While isolating controllable activities helps set goals and define roles, it does not remove uncertainty. It just shifts it. The more we focus on the results we control or outputs, the less we focus on the results we don’t control or outcomes. Many of the elements we use to reduce uncertainty, like company structures, project teams and detailed role descriptions, only reinforce the illusion that predictability brings business success.

This is especially true in the field of IT, where projects are treated as one-time, discrete initiatives for developing IT solutions. This is work that can be controlled and evaluated as either success or failure. The inherent flaw is that a project manager may perfectly deliver all planned outputs to the agreed scope, timeline and budget, but the solution may still fail to deliver business value. It’s the equivalent of saying “the surgery was successful, but the patient died.”

Contrast that with crisis management. Command centers and task forces are examples of structures that are designed to manage uncertainty. Predefined roles matter far less than initiative and speed, and outcomes matter more than outputs or process compliance. Success often hinges on collaboration and information-sharing, which requires those involved to disregard roles and responsibilities and to embrace uncertainty.

To navigate the challenges facing today’s organizations, especially those related to AI, the project-management toolset is less effective. Dealing with uncertainty is better with a management tool modeled after products, with goals and teams closer aligned to outcomes — even if it means less control, less clarity around responsibilities and if it makes personal performance evaluations harder.

This is especially imperative for organizations trying to create something novel, which requires diverse perspectives and a degree of productive friction. Rigid roles rarely generate anything new, which is why resilience must become a core capability for organizations.

Managing uncertainty

As I mentioned in Real technology transformation starts with empowering people and teams, 90% of my conversations with business leaders today begin with how they can generate revenue through AI solutions. Yet for as much as we all desire certainty, AI has introduced a tremendous amount of uncertainty into the workspace. This again underscores the importance of resilience, rather than doubling down on role clarity.

The outcomes of AI adoption and change management initiatives most IT teams are experiencing right now, regardless of how large-scale they might be or the business benefits they may produce, are inherently uncertain. Customer preferences, market shifts and emerging technologies create a constantly moving target with both unknowable and uncontrollable variables.

The key isn’t to predict every scenario, but to structure your organization to gather feedback and adjust the plan accordingly. Think of it like a GPS: when you miss a turn, it simply reroutes based on new information.

This is the backbone of product-centric thinking. It uses feedback loops from end users, analytics and market signals to continuously adjust. Real-world outcomes are prioritized over checklists, shifting the definition of success from task completion to value creation.

This frame of mind embeds resilience into teams by enabling them to manage unknowable and uncontrollable situations as they unfold, rather than trying to avoid or plan for them in advance.

Removing fear

It’s important to acknowledge that discomfort with uncertainty is closely tied to a fear of failure. “What if this doesn’t work?” will be interpreted as “who will be at fault?”

Our brains treat the discomfort of uncertainty and the fear of failure as threats, but they are not the same. As Amy Edmondson, Novartis Professor of Leadership and Management at the Harvard Business School, outlined in a paper co-authored with Emergn, “fear is the villain, not failure.”

She wrote: “In a project-driven model, failure is seen as something to avoid, based on an erroneous belief that everything will proceed as planned, leading to risk aversion … But failure can be a powerful driver of progress when it happens as the result of smart experiments within a system designed to adapt and learn.”

By testing ideas in smaller, controlled ways, teams can minimize the impact of both uncertainty and failure while maximizing continuous learning and iteration. Organizations best equipped to manage uncertainty don’t experience failure as a high-stakes event, they see it as part of change.

Uncertainty isn’t fundamentally bad — it simply challenges our instinct for control. The more IT leaders empower teams to embrace uncertainty and use it as input rather than something to resist, the better prepared they will be to think fast, act quickly and sustain momentum.

In today’s uncertain landscape, the CIO’s real job isn’t to eliminate uncertainty, but to help teams thrive within it.

This article is published as part of the Foundry Expert Contributor Network.
Want to join?