Feedback loops explain why results accelerate, stall, or swing back. There are two kinds:

1) Reinforcing loop (R) — the snowball

Amplifies change: more → even more (or less → even less).

• Example — Word of mouth: Happy customers → referrals → more sales → more resources to improve → even happier customers.
• Negative spiral: Bad reviews → fewer sales → cost cuts → worse service → more bad reviews.

Leader move: Feed what’s working (highlight wins, make sharing easy), and interrupt negative spirals fast (fix root cause, respond publicly, add support).

2) Balancing loop (B) — the thermostat

Counteracts change to keep things steady.

• Example — Workload vs. quality: Extra workload → more errors → rework slows output → pressure eases a bit (system self-corrects).
• Another view — Habits: Start a gym routine → soreness/fatigue → skip sessions → return to old baseline.

Leader move: Reduce resistance (clarify priorities, add capacity, remove blockers) so the system can stabilize at a better level, not the old one.

Actions often show results later. Training today can improve performance months from now.

Leader’s check: Set expectations and time horizons that match reality.

Rules, roles, metrics, and incentives make certain behaviors likely. People respond to the system they are in.

Leader’s check: Which rule or metric is unintentionally driving today’s results?

Deadlock: parts of the system block each other, so energy is present but nothing moves forward (e.g., speed vs. quality stalemate).

How to break it: reframe goals, add missing resources, or clarify decision rights.

Leader’s check: Where are we waiting on each other with no clear way to proceed?

Small, well-placed changes that shift the whole system (e.g., change incentives from individual wins to team wins).

Leader’s check: What one change would make most other problems easier or irrelevant?

Systems create outcomes greater than the sum of the parts (e.g., a culture of trust unlocks creativity and speed).

Leader’s check: What conditions should we create so great outcomes emerge naturally?

1. Write the problem as a pattern over time (not an event).
2. List two reinforcing loops and one balancing loop you can see.
3. Circle one rule/metric that drives the behavior.
4. Name the deadlock, if any. State how you will break it.
5. Pick one leverage point. Define a small test.
6. Run the test. Review results. Repeat.

Context: Car factories struggled with defects, waste, and slow flow. Toyota redesigned the system so problems reveal themselves and fixing them improves the whole.

Loops at work:

• Reinforcing (R): Stop the line when a defect appears → fix root cause → fewer future defects → smoother flow → more time to improve → even fewer defects.
• Balancing (B): Just-in-time supply and small batches limit overproduction and inventory spikes (keeps the system stable).
• Delay: Quality investments (training, standard work) pay off weeks/months later; reviews capture learning back into the system.

Leverage points: Empower frontline workers (andon cord). Standardized work. Visual management. Small batch sizes.

Result: Dramatic, sustained gains in quality, speed, and cost — copied globally as “Lean.”

Context: CFCs were destroying the ozone layer. Countries faced a deadlock: protect industries now vs. protect the atmosphere long term.

System design that broke the deadlock:

• Reframed goal: Phase-down over time (not overnight bans) + support for developing countries → made cooperation possible.
• Feedback loops: Regular scientific assessments and compliance reporting → transparency → trust → more countries comply → stronger results (reinforcing loop).
• Balancing loop: Targets tightened as substitutes matured, preventing backsliding.
• Delays: Even after cuts, the atmosphere needs years to heal, so monitoring continued (kept focus over time).

Leverage points: Phased targets, trade provisions, funding mechanism, and independent monitoring.

Result: Global CFC use plummeted; the ozone layer is recovering — one of history’s most successful international agreements.

Context: “Stay with a stranger” sounded risky. Without trust, growth stalls — a natural balancing loop.

How the trust system created lift:

• Balancing (problem): Low trust → few hosts/guests → few listings/reviews → little proof → trust stays low.
• Reinforcing (solution): Secure payments, insurance, verified IDs, and two-sided reviews → good stays → positive reviews → more bookings → more hosts → more choice → even better stays.
• Delays: Reviews and ratings accumulate over time; early seeding and guarantees bridged the initial trust gap.

Leverage points: Reputation system (reviews), platform protections (payments + insurance), identity verification, messaging before booking.

Result: The reinforcing trust loop unlocked global, peer-to-peer hospitality growth.