What is reinforcing feedback loop?

A reinforcing feedback loop is a closed chain where each result feeds back to amplify the original change. More leads to more, or less leads to less. It is the engine behind compounding growth, runaway decline, and winner-take-most patterns. Recognizing one tells a decision maker that small early differences will grow large.

How it works

In a reinforcing loop, a stock influences the rate that fills it. Savings earn interest, and interest raises savings, which earns more interest. The same structure works in reverse: falling revenue forces budget cuts, which reduce quality, which lowers revenue further. Reinforcing loops produce exponential curves, which start slowly and then seem to explode. They rarely continue forever, because eventually a limit or a balancing loop pushes back. Often the question is not whether growth will slow but what will slow it and when.

Reinforcing loops are neither good nor bad in themselves. The same structure that compounds savings also compounds debt. What matters is which direction the loop is turning and how long it has before something else takes over.

Example

Say a subscription business has 1,000 customers, and each month every customer brings in 0.05 new customers by recommendation, while nobody leaves. Month one adds 50 customers, giving 1,050. Month two adds 52.5, giving about 1,102. After a year the base is roughly 1,796, and after two years about 3,225. The growth rate per month is constant, but the number added each month keeps rising. This is an illustration with round numbers.

How to model it in Machinations

Draw a Pool for the stock and a Source feeding it, then use a state connection from the Pool to the Source's rate so the stock increases its own inflow. Put the rate in a Register with a formula such as a percentage of the Pool. Play the diagram with a chart of the Pool to see the curve bend upward, then change the percentage Lever to compare how sensitive the result is.