Most strategy frameworks assume the world behaves predictably. Analyse the data, identify patterns, build a plan, execute. But what happens when cause and effect are unclear, when small changes trigger outsized consequences, and when the system itself shifts as you act within it?
This is the reality of complex systems — and it demands a different approach to strategy.
Complexity thinking doesn't replace analytical rigour. It adds a layer of awareness about when traditional planning works and when it fails. The Cynefin framework, sensemaking practices, and an understanding of nonlinear dynamics give strategy leaders practical tools to navigate uncertainty without paralysis.
In this guide, I'll walk through how to recognise complexity, choose appropriate responses, and build strategy processes that match the actual nature of the challenges you face.
A complex system is one where the relationship between cause and effect is only clear in retrospect — if at all. Unlike complicated systems (where experts can analyse and predict outcomes), complex systems are characterised by emergent behaviour, interdependencies, and nonlinear dynamics.
Key characteristics of complex systems:
Emergence: Outcomes arise from interactions, not from any single component Nonlinearity: Small inputs can produce large effects (and vice versa) Feedback loops: Actions change the system, which changes the context for future actions Irreducibility: You can't understand the whole by analysing the parts in isolation
Markets, ecosystems, organisations undergoing transformation, and geopolitical environments all exhibit complex behaviour. The mistake many strategists make is treating these as merely complicated — problems that can be solved with enough analysis.
The difference matters. In complicated domains, expertise and analysis lead to correct answers. In complex domains, expertise helps you ask better questions and probe the system — but the "right answer" often only becomes visible through action and observation.
When leaders apply simple or complicated approaches to complex challenges, they often find that:
Plans become obsolete quickly — the environment shifts faster than planning cycles Predictions fail — models based on historical patterns break down Control creates fragility — rigid systems can't absorb unexpected shocks Expertise isn't enough — even domain experts can't predict emergent outcomes
The cost of misreading complexity is significant. Organisations waste resources on detailed plans that never survive contact with reality. Leaders feel blindsided by "obvious" signals they missed. Teams lose confidence in strategic processes that repeatedly fail to deliver.
Complexity thinking offers a different path: accepting uncertainty as a feature of the environment (not a failure of analysis), and building strategy processes that learn and adapt rather than predict and control.
The Cynefin framework, developed by Dave Snowden, provides a practical tool for categorising situations and choosing appropriate responses. It identifies five domains:
Characteristics: Cause and effect are obvious. Best practices exist and work reliably.
Strategic response: Sense → Categorise → Respond. Apply established solutions.
Example: Routine operational decisions with known outcomes.
Characteristics: Cause and effect are discoverable through analysis. Expert knowledge is required.
Strategic response: Sense → Analyse → Respond. Bring in expertise, study the problem, implement a solution.
Example: Engineering challenges, financial modelling, market sizing.
Characteristics: Cause and effect are only clear in retrospect. Outcomes emerge from interactions.
Strategic response: Probe → Sense → Respond. Run safe-to-fail experiments, observe patterns, amplify what works.
Example: Market entry in unfamiliar territory, cultural transformation, innovation strategy.
Characteristics: No discernible cause and effect. Crisis conditions.
Strategic response: Act → Sense → Respond. Stabilise first, then assess.
Example: Sudden market collapse, reputational crisis, supply chain disruption.
Characteristics: Unclear which domain applies. Multiple perspectives conflict.
Strategic response: Break the situation into parts and address each in its appropriate domain.
Practical application: Before choosing a strategic approach, ask: "Which domain does this challenge belong to?" The answer shapes everything from planning horizons to decision rights to how you measure success.
Start by assessing the nature of the strategic challenge. Ask:
Can we clearly define cause and effect? Do experts agree on the likely outcomes of our actions? Has this type of situation been successfully navigated before using known methods? Are we dealing with interdependent actors whose behaviour affects each other?
If the answers suggest complexity, resist the urge to force analytical certainty. Acknowledge the domain and choose methods accordingly.
In complex environments, the goal isn't to build the perfect plan — it's to learn faster than the environment changes.
Design safe-to-fail probes: small experiments that test hypotheses about what might work. Unlike pilots (which are expected to succeed), probes are designed to generate learning regardless of outcome.
Good probes are: Bounded in scope and investment Designed to reveal information about the system Reversible or adjustable based on feedback
Complex strategy requires continuous sensemaking — the ability to observe, interpret, and respond to emerging patterns.
Build feedback loops into your strategy process: Regular check-ins to assess what's changing Signals and indicators that track system behaviour Forums for discussing divergent interpretations
This is where tools like Strategy Boards become valuable — they provide a shared view where teams can track evolving assumptions, document emerging insights, and adjust direction without losing strategic coherence.
Once you observe patterns from your probes, act to amplify what's working and dampen what isn't. This is fundamentally different from "rolling out" a solution — you're nudging the system iteratively.
Questions to guide this: What's gaining traction? How can we do more of it? What's failing or stalling? Should we stop, modify, or wait? What new patterns are emerging that we didn't anticipate?
Complexity-aware strategy generates significant learning. Capture it: What hypotheses did we test? What did we observe? What did we decide, and why?
This documentation builds institutional memory and prevents teams from repeating failed experiments or abandoning promising directions prematurely.
A professional services firm wanted to expand into Southeast Asia. Traditional analysis (market sizing, competitor mapping) provided limited insight because the market was fragmented and relationships drove buying decisions in ways that weren't visible in data.
Complexity-aware approach: Rather than committing to a single market with a detailed three-year plan, they ran parallel probes in three markets — lightweight partnerships, pilot projects, and local hiring experiments. After 12 months of learning, they concentrated investment in the market where their probes had revealed the strongest fit.
A manufacturing company's digital transformation stalled because centralised planning couldn't accommodate the different readiness levels, cultures, and customer needs across business units.
Complexity-aware approach: They shifted to a portfolio of experiments — each business unit could propose and run digital initiatives within guardrails. Central strategy focused on pattern recognition: identifying what worked across contexts, creating shared infrastructure for successful approaches, and terminating experiments that showed no promise.
A SaaS company faced conflicting signals about whether to pursue enterprise or mid-market customers. Analysis was inconclusive — both segments had potential, and the product could plausibly serve either.
Complexity-aware approach: Instead of choosing based on incomplete data, they ran parallel go-to-market experiments for 90 days. The signals from actual customer behaviour (conversion, expansion, support burden) provided clarity that analysis alone couldn't.
Match your method to the domain. Don't apply complicated-domain tools (detailed forecasting, expert analysis) to complex challenges. And don't waste resources probing when the answer is actually knowable.
Make probes cheap enough to run several. In complexity, you're looking for options and learning — not betting everything on one approach.
Beware retrospective coherence. Once patterns emerge, they look obvious. This creates false confidence in prediction. Stay humble about what you "knew" versus what emerged.
Hold strategy lightly. In complex environments, commitment to a direction is good; rigid attachment to a plan is dangerous.
Create psychological safety for learning. If failed probes are punished, teams will avoid experimentation — and you'll lose your primary learning mechanism.
Use visual tools to track emergence. Complex strategy generates lots of information. Strategy Boards and similar tools help teams see patterns, track assumptions, and maintain coherence across iterations.
Complexity thinking is one component of adaptive strategy — an approach designed for environments where change is constant and prediction is limited.
Explore related guides:
What Is Adaptive Strategy? A Complete Guide — The foundational principles behind strategy that learns and evolves. Why Traditional Strategy Breaks in Fast-Moving Environments — Understanding the limitations of static planning in dynamic contexts. How to Build a 'Strategy That Learns' Using Feedback Loops — Practical implementation of learning loops in your strategy process. Stress Testing Strategy: Methods & Examples — Testing strategic choices against scenarios, including complex and uncertain futures. Strategy Under Uncertainty: A Modern Approach — Broader frameworks for developing strategy when the future is unclear.
← Return to Adaptive Strategy: A Strategy That Learns for the complete framework.
Complexity thinking becomes practical when you build it into your strategy rhythm — regular sensemaking sessions, explicit probe design, and visual tracking of what you're learning.
In Portage, Strategy Boards help you map strategic options, document assumptions, and track how your understanding evolves. The Scenario Generator lets you explore alternative futures, and stress tests reveal how your strategy holds up under different conditions — exactly the kind of probing that complex environments demand.
Map your strategy loop in Portage →
Complex systems behave differently from complicated ones — cause and effect are unclear, and expertise alone won't solve the problem. The Cynefin framework helps you match your strategic approach to the nature of the challenge. Probe-sense-respond replaces predict-plan-execute in complex domains. Safe-to-fail experiments generate learning and reduce commitment to untested assumptions. Feedback loops are essential — strategy in complexity requires continuous sensemaking and adjustment. Document your learning to build institutional memory and avoid repeating mistakes.