Systems Thinking Isn’t Enough Anymore
Why “Why You Need Systems Thinking Now” misses the deeper challenge facing leaders, boards, and organizations and tries to solve these challenges with 20th century thinking.
Introduction
In their September 2025 article for Harvard Business Review, Timan Bansal and Julian Birkinshaw “Why You Need Systems Thinking Now” revisit a familiar theme: the value of systems thinking in leadership. They urge executives to step back, look at flows and relationships, and prepare for unintended consequences. Their four-step framework is neat and appealing:
Define a future state to guide decisions.
Reframe problems through multiple lenses.
Prioritize connections over isolated parts.
Use small nudges to shift system behavior.
It’s thoughtful. It’s polished. It’s the kind of advice that feels right in a strategy session. But here’s the tension: while the framework sounds timely, it rests on assumptions that no longer hold up.
The real challenge for leaders today isn’t just seeing the system. It’s identifying the one they’re in and figuring out how to survive in systems that refuse to stay still. Furthermone, the HBR article offers a compelling what, illustrated with curated examples. But it skips the “how”:
What do you do when the system itself evolves faster than your ability to map it?
How do you lead when feedback loops misfire or nudges don’t land the way you intend?
Systems thinking still matters. Don’t get me wrong - it has its place alongside Complexity. But without tools to act in the face of uncertainty, and without a way to navigate emergence, it risks becoming a comforting narrative rather than a practical guide. That is why I found the article frustrating.
A Leadership Question
For me, the article comes down to a simple question: are you dealing with a system you can design, or one you can only navigate? If it’s the latter, maybe look at systems thinking, not as a cure-all, but rather a single lens to view the world and add complexity literacy to your tool kit, because your organization’s survival depends on it. I explore why below.
The Real Frontier
Boards aren’t managing machines; they’re navigating living systems, which requires training directors to recognize tipping points and tail risks, embedding adaptive experiments into strategy, and governing for resilience rather than perfect foresight. Is any of this even possible? Who could have predicted Microsoft’s release of an open source of Basic in 1978 would unleash an era of unprecedented innovation. No one could, only in hindsight. Systems thinking gave us the language of interconnection, but in 2025 that’s table stakes, what leaders now need is the capacity to act when uncertainty pushes back, as it always does.
From Systems Thinking to Complexity Thinking
Systems Thinking
Systems thinking was a breakthrough when it first emerged. It gave us a way to zoom out, recognize patterns, and understand how different parts of a system affect one another. As an internal auditor, I found it especially useful: it helped me make sense of how accounting systems were structured and what they were designed to achieve. It gave me a wider view than the narrow reductionist methods I had been taught.
But systems thinking can be misleading. It can tempt us into believing that if we just map carefully enough, we’ll understand the whole (I always think of Zen gardens and how it is impossible to see the entire garden from one vantage point, you have to move to see all of its features). Systems thinking assumes boundaries are clear, flows are predictable, and feedback loops can be managed. That neatness is comforting, but also false. The real world doesn’t sit politely inside our diagrams. It spills over, adapts, and surprises us.
Complexity Thinking
This is where complexity thinking takes over, in the right environment. Complexity thinking doesn’t flatter us with illusions of control. It demands that we live with uncertainty, non-linearity, and emergence. Outcomes don’t always match inputs. What worked yesterday might backfire tomorrow. In a complex system, the whole isn’t just greater than the sum of its parts, it can become something entirely different, shaped by interactions no map or org chart can capture.
So the shift isn’t about discarding systems thinking, it’s about refusing to stop there. Systems thinking got so far; it is wholly appropriate for certain situations, such as, developing an accounting process (a simple process) or designing and building a jet engine (a complicated process). Complexity thinking is another tool to help to create value. Leaders who cling only to a systems maps or their output (like heat maps in risk management) are steering with yesterday’s tools. Leaders who embrace and include complexity can learn to read the terrain as it shifts beneath their feet while following the map in their hand; they can act or move in ways to fit the terrain they face.
Some decisions can and should be made in the complicated world, where cause and effect are knowable. But many critical choices today live in the complex world, where emergence rules and adaptive leadership is required.
Complexity Literacy in Practice
Complexity literacy isn’t about abandoning systems thinking but about going deeper, relying on models rather than metaphors through simulations, stress tests, and Bayesian updating instead of clever diagrams; embracing pluralism over purity by drawing on complexity science, critical systems thinking, and soft systems methods; favoring iteration over nudging by experimenting, learning fast, and adjusting constantly; and grounding governance in aligned power and incentives rather than in good intentions or reframing exercises. The difference is visible in practice: Singapore’s COVID-19 response, with its iterative policies, data-driven shifts, and willingness to course-correct, showed adaptation in action; QatarEnergy’s LNG strategy, built on decades of scenario modeling, shock-mapping, and competitor tracking, demonstrated strategic foresight; and Toyota’s lean system, where embedded feedback and frontline empowerment mattered more than diagrams, proved how governance and culture can make resilience tangible.
The Missing “How”
The four steps outlined in the article, define a future state, reframe the problem, focus on flows, and apply nudges, sound tidy on paper. But frameworks like this tend to collapse when leaders face real-world conditions.
What does it mean to “define a desired future” when the CEO lacks credibility (think John Sculley at Apple), the board is divided, and investors punish long-term bets that will not pay off in the next quarter (unless you’re backed by private equity)? Add in the churn of political cycles, and the idea of a steady North Star feels more like wishful thinking than practical direction (and remember the North Star as an astrological phenomena is always moving and my apologies to those of you down under).
How do you “reframe problems” when shareholders and stakeholders aren’t aligned, or worse, when they’re actively working against one another? In many organizations, collaboration isn’t a default state. It’s a negotiation shaped by power plays and conflicting agendas.
And what happens when you try to “nudge” a system that pushes back (remember, that for every action there is an equal, or greater, reaction)? Real systems rarely move gracefully or as expected (if they do, count yourself lucky). They resist (most mid-level managers resist change), they compensate (draft reports that make it look like they’re complying), and sometimes they retaliate (simply by refusing to play the new game). A well-intentioned intervention can create backlash, not progress.
Strategy and Operations in a Complex World
The shortcomings of systems thinking aren’t academic, they shape how organizations build strategy and run operations.
Strategy in Motion
Traditional strategy - the way I was taught in business school - assumes stability, analyze forces, map trends, set a course, and systems thinking improved on it by surfacing interconnections, but both rest on the flawed belief that systems can be fully understood and steered. That assumption, in certain situation, no longer holds, which means leaders must move from static plans to evolving portfolios of pilots, partnerships, and minority stakes that can be scaled or abandoned as conditions shift; from prediction to scenario testing, using simulations and real-time data to refresh assumptions constantly; and from resource advantage to network advantage, recognizing that winning now depends less on owning assets and more on being embedded in adaptive ecosystems that flex with uncertainty.
Operations That Bend, Not Break
Operations optimized for efficiency may look good on paper but quickly become brittle under shock, which is why resilience depends on distributed sensing that empowers frontline teams to flag anomalies rather than follow scripts, slack and redundancy that provide buffers and backups when lean efficiency fails, modular design that allows systems to reconfigure, repurpose assets, redeploy people, and pivot supply chains, and feedback loops that don’t just collect data but trigger real action, like Toyota’s andon cord, which worked because culture made response inevitable rather than optional.
Strategy and Operations: One Adaptive Cycle
Boards often treat strategy and operations as separate tracks, but in complexity that divide collapses, strategy must be shaped by what operations sense, and operations must flex in response to strategic shifts. In supply chains, strategy isn’t just offshore versus nearshore but designing systems that can bend when climate risks or geopolitical shocks hit; in financial services, digital transformation isn’t a slogan but the operational capacity to adapt in real time to regulation, cyber threats, and shifting customers. The leaders who thrive won’t be those with the most polished plans but those who treat strategy and operations as one adaptive cycle, constantly sensing, reframing, and responding.
The Complexity Lens - Risk and Uncertainty
From a complex adaptive systems perspective, uncertainty is structural: unknown unknowns are built in, and no amount of reframing will make them predictable. Causality is fragile, as explanations make sense only in hindsight while forecasts of tipping points remain unreliable. Agents constantly adapt, with every nudge reshaping the system as people, firms, and regulators adjust in real time. Ignoring this breeds overconfidence, luring leaders into thinking tidy diagrams and reframing exercises can steer the future and when they can’t.
Why Risk and Uncertainty Aren’t the Same. And Why That Matters
The article gestures toward uncertainty by citing unintended consequences: plastics causing pollution, fracking linked to earthquakes, financial innovation leading to crisis. But, for me, the article’s framing is off: these are presented as failures of mapping, as if more careful reframing, better flow charts, or smarter nudges could have prevented surprises. That’s a comforting idea but for me its a misleading one, because what the article actually describes is risk, something that can be modeled, quantified, and priced, whereas uncertainty in the Knightian sense is different: it’s not a lack of data but a lack of knowability, and collapsing uncertainty into risk gives leaders a dangerous illusion of control.
Falling Short on Risk and Uncertainty
The glaring omission is the nature of uncertainty itself. Plastics, fracking, and financial derivatives weren’t just oversights; they were failures to navigate the uncharted. Uncertainty means we don’t know the range of outcomes, can’t fully anticipate consequences, and must accept that every intervention may trigger new, adaptive responses. To treat uncertainty as mere risk is to tame what is inherently untamable.
Why It Matters for Boards and Executives
When leaders confuse risk with uncertainty, three predictable mistakes may follow. While chasing foresight, they may assume that more mapping equals more control even as the system shifts beneath them. They may overuse nudges, forgetting that while a nudge can work in probabilistic systems, in uncertain ones they can spark backlash that is equal to or greater than the nudge itself. And they may under-invest in in adaptive capacity or resilience, by dismissing buffers and redundancy as waste when in fact these are the only shields against exposure.
A Better Way of Leading
Boards and executives do not need another reminder that the world is interconnected, they live that reality daily. What they need is a credible how: strategy that shifts from fixed roadmaps to portfolios of options that can be scaled or abandoned; operations that move from pure optimization to adaptability through slack, modular systems, and front-line feedback loops; and governance that evolves from narrow oversight to building adaptive capacity by rewarding learning, enabling rapid shifts, and aligning incentives with long-term resilience.
The Problem with Polished Case Studies
Like many business articles, “Why You Need Systems Thinking Now” leans on carefully chosen success stories. Maple Leaf Foods repositioning around sustainable food systems. An insurance cooperative baking climate risk into its offerings. A standards association reimagining its influence model. I would love to see an example of a firm that failed.
The examples are interesting, but they’re incomplete. We don’t hear about the governance fights behind the scenes. We don’t see the trade-offs made when the system resisted change. We’re left guessing how leaders translated broad ideals into action when stakeholders weren’t aligned or what it took, other than the threat of losing their job, to comply?
A handful of curated wins isn’t a method. It’s a narrative designed to fit the theory, inspire, not necessarily to instruct. Without the messy details or examples of failure, it’s impossible to know what truly worked, what truly failed, and what could be applied elsewhere (if at all).
That’s what makes the article frustrating for me - it is so one sided.
#Complexity #Leadership #Boards #Governance #Strategy #Adaptive #ComplexAdaptiveSystems
Thanks for this.
You raise important critiques about the HBR piece's "how" gap and overconfidence in control. The risk/uncertainty distinction is crucial, and you're absolutely right that adaptive systems push back against interventions in ways no diagram can predict.
But I think you're creating a false binary between systems thinking and complexity literacy. In practice, most strategic work requires both—structured analysis to understand the playing field AND adaptive capacity to navigate emergence.
As a strategy consultant, I've never seen systems thinking used as a sole prediction engine. It's an inquiry tool. When I use systems mapping with clients, the value isn't getting the "right" diagram—it's surfacing assumptions, identifying blind spots, and creating shared language for complex discussions. The framework isn't the strategy. Rather it's a conversation starter that helps my clients think through interconnections they might otherwise miss.
Your complexity science solutions sound sophisticated, but agent-based models and Bayesian updating often aren't practical given client timelines, data constraints, and cognitive bandwidth. Sometimes the "suboptimal" framework that executives can actually use beats the theoretically superior approach they can't implement.
Good strategy work already blends what you're advocating. Running scenarios, stress-test assumptions, build in optionality, and design for adaptation. Basically we need to combine structured analysis with adaptive capacity given real-world constraints.
Both of the HBR article and your response miss what actually happens in practice. We are experienced consultants who synthesize multiple approaches depending on context rather than choosing sides in academic debates.
As you indicate, much of this thinking has been around 20/30 years now - nothing new here. Complex (adaptive) systems are just a subset of systems not separate.
The generalised approach to take in a CAS environment is relatively straightforward: you amplify the areas with positive signals/outcomes and you attenuate the areas with negative signals/outcomes. And this is monitored and done continuously: there is no end-point.