Flywheel Economics
Flywheel economics describes business dynamics where multiple reinforcing loops compound into accelerating returns. Unlike a simple feedback loop (A drives B, B drives A), a flywheel chains several loops together so that momentum in one feeds momentum in all the others — and the system becomes progressively harder for competitors to replicate the longer it runs.
The canonical example is Amazon’s retail flywheel: lower prices attract more customers, more customers attract more sellers, more sellers increase selection and competition, which drives prices lower still. Each revolution of the flywheel makes the next revolution faster. Jeff Bezos reportedly sketched this on a napkin in 2001; two decades later, the flywheel has generated more than $2 trillion in market value.
Anatomy of a Flywheel
Effective flywheels share three properties. First, compounding: each cycle produces more output than the last, not merely the same amount. Second, defensibility: the accumulated momentum creates a structural advantage that new entrants cannot shortcut. Third, multi-loop reinforcement: several distinct network effects, data flywheels, and cost advantages interlock so that disrupting one loop does not stop the others.
Wright's Law is the cost engine inside many flywheels. As cumulative production grows, unit costs decline — funding lower prices, which drive more volume, which drive more cost reduction. Combined with network effects (more users make the product more valuable) and data flywheels (more usage generates better data, which improves the product), the result is a triple-helix of reinforcing dynamics that powers the most durable technology businesses.
Flywheels in Gaming and Virtual Economies
In gaming, flywheels drive the live-services model. More players generate more revenue, which funds more content updates, which attract more players. Virtual economies add a transactional flywheel: more participants make markets more liquid, which makes trading more rewarding, which attracts more participants. Roblox compounds this further with a creator economy flywheel: more players attract more developers, who build more experiences, which attract more players.
The critical insight is that flywheel strength is not about any single loop — it is about how many loops interlock. A game with only a content flywheel (more money → more content → more players) is vulnerable to a competitor with a bigger budget. A game with content, economy, social, and creator flywheels all reinforcing each other is nearly impossible to dislodge.
The Agent Flywheel
The agentic web is producing a new class of flywheel. More AI agents deployed means more demand for structured APIs and MCP servers. More tool availability makes agents more capable, which drives more agent deployment. Meanwhile, every agent interaction generates data that improves inference quality and drives costs down the Wright's Law curve — making the next generation of agents cheaper and better.
Deflationary technology supercharges all of this. As compute, bandwidth, and inference costs decline, the flywheel spins faster because each revolution requires less capital to sustain. The businesses that understand this — that price ahead of the cost curve, invest in interlocking loops rather than single advantages, and design for compounding — are the ones that build the kind of momentum that defines market eras.