Feedback Loops

Feedback Loops

The key to the feedback loop is the information it provides. You need to know whether you are moving toward your goal or away from it, and you need to know if your actions are having the intended effect.

—James Clear

Feedback loops are everywhere, making them a powerful mental model for understanding why people and systems behave the way they do. At their core, feedback loops are the engines of growth and change: they are the processes by which a system’s output influences its own input. More formally, a feedback loop is a closed chain of causal connections from a stock (some quantity that accumulates, like a bank balance or a reputation), through decisions, rules, or physical laws that depend on the level of that stock, and back again through a flow to change the stock. Without feedback, a system would simply repeat the same actions indefinitely; feedback makes systems dynamic, allowing them to adapt, amplify, or self-correct.

Think of feedback as the information communicated in response to an action. We give and receive feedback constantly, whether through formal channels like performance reviews or informal ones like body language and tone of voice. A feedback loop occurs when that information circles back and alters the system’s future behavior—much like a conversation where each reply shapes the next, making the exchange more focused and effective.

Seeing Circles, Not Lines

The key to truly grasping feedback loops is shifting from linear thinking to seeing circles of influence. In everyday language, we often describe causality in straight lines: “I am filling the glass of water,” implying a one-way chain where my hand on the faucet controls the water level. But this captures only half the story. In reality, the water level controls my hand just as much: as I perceive the gap between the current level and my desired level, I adjust the faucet. The full picture is a circle: faucet position → water flow → water level → perceived gap → hand adjustment → faucet position again. Every feedback loop tells a story, and reading that story requires tracing the recurring journey of influence, watching how the structure generates a behavioral pattern—in this case, gradually closing the faucet as the glass fills.

This shift overturns deeply ingrained assumptions. The structure causes the behavior, not a single isolated actor. My intention to fill the glass sets a system in motion, but once it’s running, I am part of the feedback process, not standing apart from it. This represents a profound shift: we are continually both influenced by and influencing our reality. It is the same shift advocated by ecologists who urge us to see ourselves as part of nature, not separate from it.

Once you start looking for feedback loops, you see them everywhere, offering insight into how incentives drive behavior. Incentives are a form of feedback: we seek actions that bring rewards or avoid punishments over various timescales. For example, if you react defensively every time a colleague offers constructive criticism, you’re creating a feedback loop that discourages honest input—ultimately depriving yourself of opportunities to grow.

The challenge is that feedback loops are so ubiquitous they can be overwhelming. How do you know which ones to pay attention to, or which to adjust to improve your outcomes? A crucial skill is learning to filter feedback. Not all feedback is useful; the faster you identify the signals that truly move you toward your goals, the quicker you progress.

Equally important is communicating feedback in a way that others can receive without defensiveness. In relationships, feedback is essential, yet many of us struggle to share honest thoughts and feelings. Much of relationship therapy involves guiding partners to voice concerns they’ve held back for years—feedback that should have been flowing freely all along. The world, too, offers us constant feedback, but we must be willing to listen and adapt rather than trying to force reality to work differently.

Because we live within many overlapping systems, feedback loops are rarely isolated. Complex systems often contain multiple interconnected loops, so a small adjustment in one area can ripple through the whole system in unexpected ways. This interconnectedness makes it difficult to appreciate the full effects of a change, but it also creates leverage points for improvement.

Balancing and Reinforcing Loops

There are two fundamental types of feedback loops: balancing (or negative) and reinforcing (or positive). Balancing loops push a system toward equilibrium, counteracting deviations. They are equilibrating or goal-seeking structures, providing both stability and resistance to change—keeping things steady but also making it hard to shift a system from its current path. A thermostat is a classic example: it senses the room’s temperature and adjusts the furnace to maintain a set point. However, for such a stock-maintaining loop to work effectively, its goal must be calibrated to compensate for any external inflows or outflows. If a window is left open, the thermostat’s feedback alone may fail to hold the desired temperature unless the set point accounts for the extra heat loss; otherwise, the system will consistently undershoot or overshoot.

Reinforcing loops, by contrast, amplify what is happening, driving growth or decline. They are self-perpetuating and can lead to exponential growth—like a fashion trend that explodes in popularity before burning out—or to runaway collapses, such as a bank run where each withdrawal deepens panic and triggers more withdrawals. The compound disadvantages that trap people in poverty are another example of a reinforcing loop that spirals downward without outside intervention. Breaking out of reinforcing loops often requires a significant shift in conditions or an external disruption.

The Problem of Delays, Responsibility, and Language

One of the biggest challenges is that feedback is rarely immediate, especially in complex systems. Delays between an action and its visible effect make it hard to establish cause and effect. In our personal lives, this can lead to maladaptive behaviors: something like eating junk food delivers an instant hit of pleasure, but the negative feedback—from sluggishness to long-term health issues like diabetes—arrives much later. When feedback is delayed, we risk overvaluing the short-term signal and missing the more important long-term message. Moreover, even if feedback were perfectly instantaneous, it can still only influence future behavior—it can never retroactively correct the actions that generated it. This inherent forward-looking nature means that any system always operates with some lag, however slight, and we must learn to act with that in mind.

The speed and quality of feedback also matter. Faster, more accurate feedback allows for rapid iteration and improvement, much like a driver quickly correcting course. However, if feedback comes too fast or is too strong, it can cause a system to surge out of control—imagine a new driver stomping on the gas and brake. The art of using feedback loops wisely involves not just noticing them but also managing the tempo of your response, monitoring how changes ripple through the system, and staying open to revision.

Adopting a feedback perspective also carries ethical weight. Linear thinking always points to a simple locus of responsibility: when things go wrong, there is either blame (“they did it”) or guilt (“I did it”). But from a feedback viewpoint, everyone shares responsibility for problems generated by a system. This doesn’t mean everyone can exert equal leverage to change it, but it does mean that the search for scapegoats—an alluring pastime in many cultures—is a dead end. The very language we use reinforces linear habits. Describing even a simple feedback system like filling a glass of water becomes awkward in everyday English. If it’s this clumsy to articulate a single loop, imagine the difficulty of capturing the multiple, interlocking feedback processes in an organization, an economy, or an ecosystem using only our conventional linear vocabulary.

Understanding feedback loops reveals a deeper pattern: systems with similar feedback structures tend to exhibit similar dynamic behaviors, regardless of their specific details. A reinforcing loop of distrust in a relationship can spiral in the same way as a reinforcing loop of debt accumulation, both racing toward collapse unless interrupted. This universality makes fluency in feedback loops a powerful tool for diagnosing and improving the systems we live and work within.

Ultimately, feedback loops are what make systems dynamic. Without them, a system would simply repeat the same pattern over and over. Understand them, respect them, and use them wisely.