A stock is anything that accumulates: money in an account, water in a tub, trust in a team, quality in a codebase. You can measure it at a moment in time. A flow is a rate — the filling and draining that changes the stock over time.

That's the whole model. What makes it worth learning is what it predicts about how systems behave.

Fig. 1The level is the stock. The faucet and the drain are flows. Nothing else in the system exists.

Inertia is the insight

Stocks give systems inertia. The level moves slowly because flows take time to fill or drain it — long after you fix the faucet, the tub is still mostly where it was. This is why climate change is hard to reverse, why getting fit takes months, why housing shortages persist for decades, and why technical debt piles up quietly until it stops feeling quiet.

“You can't fix a leaky bucket by pouring faster.”

When a system frustrates you, the reflex is to push harder on the inflow: more marketing, more hours, more hiring. But if the drain is open, more pouring just raises the bill. The leverage is usually at the outflow nobody is watching.

A worked example: technical debt

Treat code quality as the stock. Refactoring and documentation fill it; shortcuts and features-in-a-hurry drain it.

Fig. 2Velocity is a flow. Quality is the stock. A sprint tells you nothing about the level.

Teams that never refactor aren't unlucky — their outflow simply exceeds their inflow, and the stock obeys the arithmetic. The decline doesn't show up in any single week, because no single week moves a stock much. Then one day the codebase feels immovable, and everyone calls it mysterious.

The pattern is everywhere

Once you name the stock, the flows become obvious — and so does where the leverage is.

SystemStockFilled byDrained by
User growthActive usersNew signupsChurn, inactivity
Knowledge baseTeam knowledge, runbooksLearning, writing docsDepartures, staleness
InfrastructureAvailable capacityProvisioningUsage growth, decommissioning
TrustBrand trust, user confidenceGood experiences, transparencyBugs, outages, breaches
ExpertiseWhat you actually knowDeliberate practiceForgetting, obsolescence
EnergyCapacity to do hard thingsRest, recovery, joyStress, context switching
RelationshipsDepth of connectionsInvesting time in peopleNeglect, moving, changed contexts
AttentionAbility to concentrateDeep work, single-taskingDistraction, notifications
RunwaySavings, investmentsIncome, returnsExpenses, lifestyle inflation

Two rows deserve a second look. Trust builds slowly and drains fast — one breach can empty years of stock, so the asymmetry of the flow rates matters as much as the level. And expertise can't be crammed: that stock only fills through consistent inflow, which is why reading about something once does approximately nothing.

When stocks feed their own flows

The compounding cases are the interesting ones. Sometimes a rising stock increases its own inflow rate — and the system starts feeding itself.

Network effects

Past critical mass, each new user makes the product more valuable, which raises the signup rate. The stock amplifies its own inflow.

Skill compounding

The more you know about a domain, the faster you learn adjacent skills. Expertise makes acquiring expertise cheaper.

Good architecture

When features are easy to add, you're less tempted to take shortcuts — which keeps features easy to add.

Documentation quality

Good docs bring in people who improve the docs. A self-reinforcing stock of shared understanding.

What this changes

You won't feel it draining

Feedback from a stock is delayed. You can run a deficit on sleep, relationships, code quality, or trust for a long time before the level hits zero and everything breaks at once. The absence of pain is not evidence the stock is fine.

Optimize the level, not the rate

You're optimizing for the long-term stock, not this week's flow. That's the honest argument for sustainable pace: heroic sprints spike a flow while quietly draining the stocks — energy, quality, trust — that all future flow depends on.

Invest in stocks, not flows

Working harder pushes on a flow, once. Building — better tools, better docs, better architecture, better habits — is a one-time investment that permanently raises an inflow rate. The real leverage compounds when loops stack, and one stock's surplus feeds another's inflow:

Source This mental model comes from Donella Meadows' work on system dynamics — Thinking in Systems is the canonical read. It's popular in tech and product circles because it explains why some problems stay stubbornly persistent no matter how hard anyone intervenes.