Mental Models

Feedback Loops

A feedback loop exists when the output of a system influences its own future input, either reinforcing or dampening the original signal.

Understanding feedback loops reveals why small changes sometimes explode into large effects and why some problems self-correct while others spiral.

Core question

Is the output of this system reinforcing or correcting the behavior that produced it?

Example

A product with good onboarding generates positive reviews, which attract more users, which fund better onboarding. This positive feedback loop compounds growth.

How to apply

1.Map the causal chain from action to outcome and back.
2.Label each loop as reinforcing (amplifying) or balancing (stabilizing).
3.Identify delay between action and feedback — long delays hide loops.
4.Design interventions at the highest-leverage point in the loop.

Pitfalls

·Ignoring time delays that make feedback loops invisible until they dominate.
·Assuming all reinforcing loops are positive — they amplify bad dynamics too.
·Intervening at low-leverage points and wondering why the system does not change.

Questions

What is a negative feedback loop example?

A thermostat: when temperature rises above the setpoint, cooling activates and brings it back down. The output corrects the input.

How do feedback loops relate to product growth?

Viral loops, network effects, and retention flywheels are all reinforcing feedback loops that compound adoption over time.

Compounding
Small, consistent gains accumulate into outsized results over time.
Systems Thinking
Understand behavior by examining the whole system, not just the parts.
Network Effects
A product becomes more valuable as more people use it.