_In control theory, some systems move the wrong way first. Push right, they go left — then correct. The danger is fighting the initial movement._ --- ## The physics You walk into a turnaround. Costs are too high, quality is inconsistent, customers are churning. You make changes. Three months later, the numbers look worse. This is the moment that separates leaders who transform businesses from those who oscillate between initiatives until someone replaces them. Engineers call it a right-half-plane zero. When you apply input that should move the system up, it moves down first — then corrects. Push an aircraft's tail down to climb, and the plane briefly drops before rising. Turn a bicycle right, and you countersteer left first. The critical insight: if you see the initial wrong-way movement and overcorrect, you amplify the problem. The system swings harder the other way, you correct again, oscillations grow. The Chernobyl reactor exhibited this dynamic in its final minutes. The correct response to inverse response is to hold steady and let the system work through it. --- ## What this looks like in a business A new leader arrives and makes necessary changes. Stops unprofitable work — revenue drops before costs fall. Invests in fixing root causes — expenses rise before quality improves. Resets customer expectations — satisfaction scores dip before trust rebuilds. Raises standards — short-term output falls before capability grows. Every one of these is correct. And every one looks identical to "new leader making things worse." The board sees declining metrics. The team feels the strain of change without visible payoff. Customers experience disruption. Everyone is watching the chart move the wrong direction, asking: does this person know what they're doing, or are they destroying value? --- ## The credibility problem You can't avoid the dip. What you can control is whether stakeholders trust you through it. Predict it in advance. Before the numbers turn, tell people what's coming and why. "We'll see margins compress in Q1 as we exit low-value contracts. By Q3, the remaining book will be healthier." This transforms the dip from evidence of failure into confirmation of the plan. Define leading indicators. Lagging metrics — revenue, profit, NPS — will look bad while the system corrects. But leading indicators should move first: pipeline quality, employee engagement, defect rates, customer retention in target segments. These show the change is working before the headline numbers reflect it. Set time horizons. "If we don't see X by month six, we'll revisit the approach." This creates a decision point — a moment where you and the board can distinguish inverse response from genuine failure. Without it, you're asking for indefinite faith. --- ## The opposite failures The leaders who fail turnarounds usually fail one of two ways. Panic and reverse. They see the dip, lose nerve, and undo the changes. Now you've incurred the transition costs twice — once going in, once coming out — without getting the benefit. Worse, you've taught the organisation that change initiatives don't stick. Fail to prepare stakeholders. They make the right changes but don't prepare the board for the dip. When the numbers drop, the board loses confidence and pulls the plug before the system can correct. The leader was right about the physics but wrong about the politics. Both look like failed transformation. Neither was actually a failure of strategy. --- ## Distinguishing dip from failure The hard part: inverse response and genuine failure look the same at first. Both show the numbers moving wrong. The difference is in the leading indicators. In inverse response, the system is restructuring — early signals of health appear even while lagging metrics decline. In genuine failure, everything deteriorates together. This is why prediction matters. If you understand the dynamics of your system, you can predict what inverse response should look like. When reality matches that prediction, hold steady. When it doesn't, course correct. The discipline is neither blind faith nor reactive panic. It's having a model, watching for confirmation, and knowing in advance what would change your mind. --- **See also:** [[Ideas/Systems Thinking|Systems Thinking]] · [[Ideas/OODA Loop|OODA Loop]]