Systems Thinking
Systems Thinking
What is systems thinking?
- Seeing the connected whole, not just the parts
- Avoiding the trap of local optimization
- Finding root causes and leverage points
What systems thinking actually changes is the question asked in the meeting room. Where people used to ask 'Who dropped the ball this time?', they start asking 'What structure keeps producing this result?' If the same incident recurs even after you replace the person in charge or issue new guidelines, the system, not the people, is producing it. The moment you move the problem from people to structure, the nature of the fix changes too. Instead of reprimands and pledges to prevent recurrence, the focus shifts to reworking the system itself—KPIs, budget allocation, and approval routes.
The tools used in practice are simpler than you might expect. You draw a causal loop diagram (CLD) linking elements with arrows, then distinguish two kinds of loops: reinforcing loops that feed themselves (revenue grows, so you advertise more, and revenue grows again) and balancing loops that hold things near a target (inventory rises, so you cut production). Mark the delays between cause and effect and the picture is complete. It works differently from MECE or 5 Why. Where those two are linear decomposition—splitting things apart and tracing them backward—systems thinking reconnects the pieces to see the cycle.
The most valuable piece of information is the length of the delay. A new hire shows up in productivity only 3–6 months after joining, a quality investment appears as fewer claims 6–12 months later, and organizational culture interventions can take more than a year. If you don’t know these lags, you reverse the policy just before it starts to pay off. So when designing an intervention, first write down “what should be visible, and when,” and get executives to agree to hold off on judgment until that point.
Teams most often go wrong by feeling they have to map everything. If you fill a whiteboard with arrows, spend three months, and still can’t pick one place to act, that isn’t analysis—it’s avoidance. The opposite misunderstanding is just as risky. When “it’s the structure, not the people” becomes an excuse, no one takes accountability for decisions. Some teams also use the fear of local optimization to block even small front-line improvements, but the moment “see the whole” turns into “change nothing”, the organization keeps analyzing and misses the market.
John KimInstructor John Kim, author of "B2B Marketing Design" and CEO of Nanum Consulting
Sungmin JooExpert in assessment-based coaching leadership, entrepreneurship education, and design thinkingPopularized by Peter Senge in The Fifth Discipline (1990) as a core principle of the learning organization.
Consider a hypothetical case. Company A makes home appliance components, with 420 employees and annual revenue of KRW 120 billion (≈ US$89,780,000). Over the past six months its on-time delivery rate fell to 82%, and two major clients sent penalty notices. Yet over the same period, production line utilization hit a record high of 92%.
- After the head of production reported, “We ran the line to target,” the sales director asked, “Then why are customers waiting three more weeks? Those two numbers have never improved together.”
- When the planning team lead drew a causal map over four weeks, a loop emerged. The utilization KPI encouraged large batches, inventory piled up, urgent orders cut in, setups increased, and deliveries slipped again—a self-reinforcing loop.
- They also looked at the time lag. A preventive maintenance budget cut two years earlier had pushed equipment failures from 4 to 11 per month eight months later, and at the time no one connected the two events.
- They pulled only one lever. They changed the production division’s top KPI from utilization to on-time delivery rate and order lead time, cut the average batch size in half, and required plant-manager approval for urgent orders.
- Six months later, on-time delivery had risen from 82% to 95%, and inventory turnover from 4.1 to 6.3 times. Utilization fell to 87%, but monthly overtime costs dropped from KRW 32 million (≈ US$23,900) to KRW 19 (≈ US$1) million, so no one objected.