Thinking Tools4.2 Expansive Idea FormationAlso known as Twelve Leverage Points, Places to Intervene in a System
Developed by Donella H. Meadows (Leverage Points: Places to Intervene in a System, 1999).
Ranks the places to intervene in a system, from adjusting numbers up through rules, goals, and mindsets, showing which changes matter most.
Eight fixes for rising warehouse injuries, placed on the twelve leverage points
The twelve points run from the weakest place to intervene, bottom left, to the strongest, top right. The bonus the company already tried is placed; place each other fix by what it changes, and see how far up the ladder it reaches. Illustrative.
Bigger safe-shift bonus. Already tried: the bonus for an injury-free shift was raised twice.Not placed yet.Bigger safe-shift bonus. Already tried: the bonus for an injury-free shift was raised twice.12 Numbers: Rates, fines, targets, and budgets. Easy to change and argued over most, they rarely change how the system behaves.It changes a number and nothing else. The crews still rush, so the injuries kept rising.Bigger safe-shift bonus. Already tried: the bonus for an injury-free shift was raised twice.11 Buffers: The stocks that absorb swings, sized against their flows. A bigger buffer steadies the system, but it costs money and is slow to build.Bigger safe-shift bonus. Already tried: the bonus for an injury-free shift was raised twice.10 Layout: How the stocks and flows are physically laid out. Strong when designed in, slow and costly to rebuild later.Bigger safe-shift bonus. Already tried: the bonus for an injury-free shift was raised twice.9 Delays: The lag between a change and the signal that reports it. A lag long against the pace of the system makes people overshoot.Bigger safe-shift bonus. Already tried: the bonus for an injury-free shift was raised twice.8 Balancing: Balancing loops: the ones that pull the system back toward its goal, and how strong they are against what they correct.Bigger safe-shift bonus. Already tried: the bonus for an injury-free shift was raised twice.7 Reinforcing: Reinforcing loops, the ones that feed on themselves. Slowing a runaway loop usually does more than adding brakes against it.Bigger safe-shift bonus. Already tried: the bonus for an injury-free shift was raised twice.6 Information: Who sees which signal, and when. Showing the people who act a signal they lacked changes behavior without a new rule.Bigger safe-shift bonus. Already tried: the bonus for an injury-free shift was raised twice.5 Rules: Incentives, penalties, and constraints, and who sets them. Change the rules and behavior follows.Bigger safe-shift bonus. Already tried: the bonus for an injury-free shift was raised twice.4 Redesign: The power of the people in the system to redesign its structure themselves: add, change, or drop parts of it.Bigger safe-shift bonus. Already tried: the bonus for an injury-free shift was raised twice.3 Goals: What the system is for. Everything below is arranged to serve the goal, so changing the goal rearranges it.Bigger safe-shift bonus. Already tried: the bonus for an injury-free shift was raised twice.2 Mindset: The shared beliefs the goals, rules, and structure come from. Rarely shifted, and decisive when they are.Bigger safe-shift bonus. Already tried: the bonus for an injury-free shift was raised twice.1 Transcend: Holding no mindset as final, and choosing the one that serves the purpose at hand.A spare picker per shift. One extra person on each shift to cover the rush hours.Not placed yet.A spare picker per shift. One extra person on each shift to cover the rush hours.12 Numbers: Rates, fines, targets, and budgets. Easy to change and argued over most, they rarely change how the system behaves.A spare picker per shift. One extra person on each shift to cover the rush hours.11 Buffers: The stocks that absorb swings, sized against their flows. A bigger buffer steadies the system, but it costs money and is slow to build.A buffer against the peaks. It helps, but it adds cost every shift and leaves the rush itself in place.A spare picker per shift. One extra person on each shift to cover the rush hours.10 Layout: How the stocks and flows are physically laid out. Strong when designed in, slow and costly to rebuild later.A spare picker per shift. One extra person on each shift to cover the rush hours.9 Delays: The lag between a change and the signal that reports it. A lag long against the pace of the system makes people overshoot.A spare picker per shift. One extra person on each shift to cover the rush hours.8 Balancing: Balancing loops: the ones that pull the system back toward its goal, and how strong they are against what they correct.A spare picker per shift. One extra person on each shift to cover the rush hours.7 Reinforcing: Reinforcing loops, the ones that feed on themselves. Slowing a runaway loop usually does more than adding brakes against it.A spare picker per shift. One extra person on each shift to cover the rush hours.6 Information: Who sees which signal, and when. Showing the people who act a signal they lacked changes behavior without a new rule.A spare picker per shift. One extra person on each shift to cover the rush hours.5 Rules: Incentives, penalties, and constraints, and who sets them. Change the rules and behavior follows.A spare picker per shift. One extra person on each shift to cover the rush hours.4 Redesign: The power of the people in the system to redesign its structure themselves: add, change, or drop parts of it.A spare picker per shift. One extra person on each shift to cover the rush hours.3 Goals: What the system is for. Everything below is arranged to serve the goal, so changing the goal rearranges it.A spare picker per shift. One extra person on each shift to cover the rush hours.2 Mindset: The shared beliefs the goals, rules, and structure come from. Rarely shifted, and decisive when they are.A spare picker per shift. One extra person on each shift to cover the rush hours.1 Transcend: Holding no mindset as final, and choosing the one that serves the purpose at hand.Same-day near-miss notes. Near misses now reach the site manager a month later, in a summary.Not placed yet.Same-day near-miss notes. Near misses now reach the site manager a month later, in a summary.12 Numbers: Rates, fines, targets, and budgets. Easy to change and argued over most, they rarely change how the system behaves.Same-day near-miss notes. Near misses now reach the site manager a month later, in a summary.11 Buffers: The stocks that absorb swings, sized against their flows. A bigger buffer steadies the system, but it costs money and is slow to build.Same-day near-miss notes. Near misses now reach the site manager a month later, in a summary.10 Layout: How the stocks and flows are physically laid out. Strong when designed in, slow and costly to rebuild later.Same-day near-miss notes. Near misses now reach the site manager a month later, in a summary.9 Delays: The lag between a change and the signal that reports it. A lag long against the pace of the system makes people overshoot.It shortens a delay from a month to a day, so the manager can act while the cause is still on the floor.Same-day near-miss notes. Near misses now reach the site manager a month later, in a summary.8 Balancing: Balancing loops: the ones that pull the system back toward its goal, and how strong they are against what they correct.Same-day near-miss notes. Near misses now reach the site manager a month later, in a summary.7 Reinforcing: Reinforcing loops, the ones that feed on themselves. Slowing a runaway loop usually does more than adding brakes against it.Same-day near-miss notes. Near misses now reach the site manager a month later, in a summary.6 Information: Who sees which signal, and when. Showing the people who act a signal they lacked changes behavior without a new rule.Same-day near-miss notes. Near misses now reach the site manager a month later, in a summary.5 Rules: Incentives, penalties, and constraints, and who sets them. Change the rules and behavior follows.Same-day near-miss notes. Near misses now reach the site manager a month later, in a summary.4 Redesign: The power of the people in the system to redesign its structure themselves: add, change, or drop parts of it.Same-day near-miss notes. Near misses now reach the site manager a month later, in a summary.3 Goals: What the system is for. Everything below is arranged to serve the goal, so changing the goal rearranges it.Same-day near-miss notes. Near misses now reach the site manager a month later, in a summary.2 Mindset: The shared beliefs the goals, rules, and structure come from. Rarely shifted, and decisive when they are.Same-day near-miss notes. Near misses now reach the site manager a month later, in a summary.1 Transcend: Holding no mindset as final, and choosing the one that serves the purpose at hand.Cap the pick-rate bonus. Crews earn more the faster they pick, so faster crews earn more and push harder.Not placed yet.Cap the pick-rate bonus. Crews earn more the faster they pick, so faster crews earn more and push harder.12 Numbers: Rates, fines, targets, and budgets. Easy to change and argued over most, they rarely change how the system behaves.Cap the pick-rate bonus. Crews earn more the faster they pick, so faster crews earn more and push harder.11 Buffers: The stocks that absorb swings, sized against their flows. A bigger buffer steadies the system, but it costs money and is slow to build.Cap the pick-rate bonus. Crews earn more the faster they pick, so faster crews earn more and push harder.10 Layout: How the stocks and flows are physically laid out. Strong when designed in, slow and costly to rebuild later.Cap the pick-rate bonus. Crews earn more the faster they pick, so faster crews earn more and push harder.9 Delays: The lag between a change and the signal that reports it. A lag long against the pace of the system makes people overshoot.Cap the pick-rate bonus. Crews earn more the faster they pick, so faster crews earn more and push harder.8 Balancing: Balancing loops: the ones that pull the system back toward its goal, and how strong they are against what they correct.Cap the pick-rate bonus. Crews earn more the faster they pick, so faster crews earn more and push harder.7 Reinforcing: Reinforcing loops, the ones that feed on themselves. Slowing a runaway loop usually does more than adding brakes against it.It slows the loop that pays people to rush, which no safety bonus could outweigh.Cap the pick-rate bonus. Crews earn more the faster they pick, so faster crews earn more and push harder.6 Information: Who sees which signal, and when. Showing the people who act a signal they lacked changes behavior without a new rule.Cap the pick-rate bonus. Crews earn more the faster they pick, so faster crews earn more and push harder.5 Rules: Incentives, penalties, and constraints, and who sets them. Change the rules and behavior follows.Cap the pick-rate bonus. Crews earn more the faster they pick, so faster crews earn more and push harder.4 Redesign: The power of the people in the system to redesign its structure themselves: add, change, or drop parts of it.Cap the pick-rate bonus. Crews earn more the faster they pick, so faster crews earn more and push harder.3 Goals: What the system is for. Everything below is arranged to serve the goal, so changing the goal rearranges it.Cap the pick-rate bonus. Crews earn more the faster they pick, so faster crews earn more and push harder.2 Mindset: The shared beliefs the goals, rules, and structure come from. Rarely shifted, and decisive when they are.Cap the pick-rate bonus. Crews earn more the faster they pick, so faster crews earn more and push harder.1 Transcend: Holding no mindset as final, and choosing the one that serves the purpose at hand.Near misses on screens. The screens show picks per hour, nothing about safety.Not placed yet.Near misses on screens. The screens show picks per hour, nothing about safety.12 Numbers: Rates, fines, targets, and budgets. Easy to change and argued over most, they rarely change how the system behaves.Near misses on screens. The screens show picks per hour, nothing about safety.11 Buffers: The stocks that absorb swings, sized against their flows. A bigger buffer steadies the system, but it costs money and is slow to build.Near misses on screens. The screens show picks per hour, nothing about safety.10 Layout: How the stocks and flows are physically laid out. Strong when designed in, slow and costly to rebuild later.Near misses on screens. The screens show picks per hour, nothing about safety.9 Delays: The lag between a change and the signal that reports it. A lag long against the pace of the system makes people overshoot.Near misses on screens. The screens show picks per hour, nothing about safety.8 Balancing: Balancing loops: the ones that pull the system back toward its goal, and how strong they are against what they correct.Near misses on screens. The screens show picks per hour, nothing about safety.7 Reinforcing: Reinforcing loops, the ones that feed on themselves. Slowing a runaway loop usually does more than adding brakes against it.Near misses on screens. The screens show picks per hour, nothing about safety.6 Information: Who sees which signal, and when. Showing the people who act a signal they lacked changes behavior without a new rule.It puts a signal the crews never saw in front of the people who can act on it.Near misses on screens. The screens show picks per hour, nothing about safety.5 Rules: Incentives, penalties, and constraints, and who sets them. Change the rules and behavior follows.Near misses on screens. The screens show picks per hour, nothing about safety.4 Redesign: The power of the people in the system to redesign its structure themselves: add, change, or drop parts of it.Near misses on screens. The screens show picks per hour, nothing about safety.3 Goals: What the system is for. Everything below is arranged to serve the goal, so changing the goal rearranges it.Near misses on screens. The screens show picks per hour, nothing about safety.2 Mindset: The shared beliefs the goals, rules, and structure come from. Rarely shifted, and decisive when they are.Near misses on screens. The screens show picks per hour, nothing about safety.1 Transcend: Holding no mindset as final, and choosing the one that serves the purpose at hand.Stop line, keep credit. Today a picker who stops the line loses that shift's pick credit.Not placed yet.Stop line, keep credit. Today a picker who stops the line loses that shift's pick credit.12 Numbers: Rates, fines, targets, and budgets. Easy to change and argued over most, they rarely change how the system behaves.Stop line, keep credit. Today a picker who stops the line loses that shift's pick credit.11 Buffers: The stocks that absorb swings, sized against their flows. A bigger buffer steadies the system, but it costs money and is slow to build.Stop line, keep credit. Today a picker who stops the line loses that shift's pick credit.10 Layout: How the stocks and flows are physically laid out. Strong when designed in, slow and costly to rebuild later.Stop line, keep credit. Today a picker who stops the line loses that shift's pick credit.9 Delays: The lag between a change and the signal that reports it. A lag long against the pace of the system makes people overshoot.Stop line, keep credit. Today a picker who stops the line loses that shift's pick credit.8 Balancing: Balancing loops: the ones that pull the system back toward its goal, and how strong they are against what they correct.Stop line, keep credit. Today a picker who stops the line loses that shift's pick credit.7 Reinforcing: Reinforcing loops, the ones that feed on themselves. Slowing a runaway loop usually does more than adding brakes against it.Stop line, keep credit. Today a picker who stops the line loses that shift's pick credit.6 Information: Who sees which signal, and when. Showing the people who act a signal they lacked changes behavior without a new rule.Stop line, keep credit. Today a picker who stops the line loses that shift's pick credit.5 Rules: Incentives, penalties, and constraints, and who sets them. Change the rules and behavior follows.It removes the penalty for doing the safe thing, so stopping the line stops costing money.Stop line, keep credit. Today a picker who stops the line loses that shift's pick credit.4 Redesign: The power of the people in the system to redesign its structure themselves: add, change, or drop parts of it.Stop line, keep credit. Today a picker who stops the line loses that shift's pick credit.3 Goals: What the system is for. Everything below is arranged to serve the goal, so changing the goal rearranges it.Stop line, keep credit. Today a picker who stops the line loses that shift's pick credit.2 Mindset: The shared beliefs the goals, rules, and structure come from. Rarely shifted, and decisive when they are.Stop line, keep credit. Today a picker who stops the line loses that shift's pick credit.1 Transcend: Holding no mindset as final, and choosing the one that serves the purpose at hand.Goal: injury-free picks. The posted goal is picks per hour, site against site.Not placed yet.Goal: injury-free picks. The posted goal is picks per hour, site against site.12 Numbers: Rates, fines, targets, and budgets. Easy to change and argued over most, they rarely change how the system behaves.Goal: injury-free picks. The posted goal is picks per hour, site against site.11 Buffers: The stocks that absorb swings, sized against their flows. A bigger buffer steadies the system, but it costs money and is slow to build.Goal: injury-free picks. The posted goal is picks per hour, site against site.10 Layout: How the stocks and flows are physically laid out. Strong when designed in, slow and costly to rebuild later.Goal: injury-free picks. The posted goal is picks per hour, site against site.9 Delays: The lag between a change and the signal that reports it. A lag long against the pace of the system makes people overshoot.Goal: injury-free picks. The posted goal is picks per hour, site against site.8 Balancing: Balancing loops: the ones that pull the system back toward its goal, and how strong they are against what they correct.Goal: injury-free picks. The posted goal is picks per hour, site against site.7 Reinforcing: Reinforcing loops, the ones that feed on themselves. Slowing a runaway loop usually does more than adding brakes against it.Goal: injury-free picks. The posted goal is picks per hour, site against site.6 Information: Who sees which signal, and when. Showing the people who act a signal they lacked changes behavior without a new rule.Goal: injury-free picks. The posted goal is picks per hour, site against site.5 Rules: Incentives, penalties, and constraints, and who sets them. Change the rules and behavior follows.Goal: injury-free picks. The posted goal is picks per hour, site against site.4 Redesign: The power of the people in the system to redesign its structure themselves: add, change, or drop parts of it.Goal: injury-free picks. The posted goal is picks per hour, site against site.3 Goals: What the system is for. Everything below is arranged to serve the goal, so changing the goal rearranges it.It changes what the system is for, and every rule and screen gets rearranged to serve it. Expect a fight.Goal: injury-free picks. The posted goal is picks per hour, site against site.2 Mindset: The shared beliefs the goals, rules, and structure come from. Rarely shifted, and decisive when they are.Goal: injury-free picks. The posted goal is picks per hour, site against site.1 Transcend: Holding no mindset as final, and choosing the one that serves the purpose at hand.Safe picks are fast. Supervisors say injuries are the price of hitting volume.Not placed yet.Safe picks are fast. Supervisors say injuries are the price of hitting volume.12 Numbers: Rates, fines, targets, and budgets. Easy to change and argued over most, they rarely change how the system behaves.Safe picks are fast. Supervisors say injuries are the price of hitting volume.11 Buffers: The stocks that absorb swings, sized against their flows. A bigger buffer steadies the system, but it costs money and is slow to build.Safe picks are fast. Supervisors say injuries are the price of hitting volume.10 Layout: How the stocks and flows are physically laid out. Strong when designed in, slow and costly to rebuild later.Safe picks are fast. Supervisors say injuries are the price of hitting volume.9 Delays: The lag between a change and the signal that reports it. A lag long against the pace of the system makes people overshoot.Safe picks are fast. Supervisors say injuries are the price of hitting volume.8 Balancing: Balancing loops: the ones that pull the system back toward its goal, and how strong they are against what they correct.Safe picks are fast. Supervisors say injuries are the price of hitting volume.7 Reinforcing: Reinforcing loops, the ones that feed on themselves. Slowing a runaway loop usually does more than adding brakes against it.Safe picks are fast. Supervisors say injuries are the price of hitting volume.6 Information: Who sees which signal, and when. Showing the people who act a signal they lacked changes behavior without a new rule.Safe picks are fast. Supervisors say injuries are the price of hitting volume.5 Rules: Incentives, penalties, and constraints, and who sets them. Change the rules and behavior follows.Safe picks are fast. Supervisors say injuries are the price of hitting volume.4 Redesign: The power of the people in the system to redesign its structure themselves: add, change, or drop parts of it.Safe picks are fast. Supervisors say injuries are the price of hitting volume.3 Goals: What the system is for. Everything below is arranged to serve the goal, so changing the goal rearranges it.Safe picks are fast. Supervisors say injuries are the price of hitting volume.2 Mindset: The shared beliefs the goals, rules, and structure come from. Rarely shifted, and decisive when they are.It challenges the belief the goal and the bonus came from. Slow to shift, and nothing else holds without it.Safe picks are fast. Supervisors say injuries are the price of hitting volume.1 Transcend: Holding no mindset as final, and choosing the one that serves the purpose at hand.
Your sort matched 0 of 8.Your sort matched 1 of 8.Your sort matched 2 of 8.Your sort matched 3 of 8.Your sort matched 4 of 8.Your sort matched 5 of 8.Your sort matched 6 of 8.Your sort matched 7 of 8.Your sort matched 8 of 8.
12 Numbers
Rates, fines, targets, and budgets. Easy to change and argued over most, they rarely change how the system behaves.
11 Buffers
The stocks that absorb swings, sized against their flows. A bigger buffer steadies the system, but it costs money and is slow to build.
10 Layout
How the stocks and flows are physically laid out. Strong when designed in, slow and costly to rebuild later.
9 Delays
The lag between a change and the signal that reports it. A lag long against the pace of the system makes people overshoot.
8 Balancing
Balancing loops: the ones that pull the system back toward its goal, and how strong they are against what they correct.
7 Reinforcing
Reinforcing loops, the ones that feed on themselves. Slowing a runaway loop usually does more than adding brakes against it.
6 Information
Who sees which signal, and when. Showing the people who act a signal they lacked changes behavior without a new rule.
5 Rules
Incentives, penalties, and constraints, and who sets them. Change the rules and behavior follows.
4 Redesign
The power of the people in the system to redesign its structure themselves: add, change, or drop parts of it.
3 Goals
What the system is for. Everything below is arranged to serve the goal, so changing the goal rearranges it.
2 Mindset
The shared beliefs the goals, rules, and structure come from. Rarely shifted, and decisive when they are.
1 Transcend
Holding no mindset as final, and choosing the one that serves the purpose at hand.
The sort
Bigger safe-shift bonus: 12 Numbers. It changes a number and nothing else. The crews still rush, so the injuries kept rising.
A spare picker per shift: 11 Buffers. A buffer against the peaks. It helps, but it adds cost every shift and leaves the rush itself in place.
Same-day near-miss notes: 9 Delays. It shortens a delay from a month to a day, so the manager can act while the cause is still on the floor.
Cap the pick-rate bonus: 7 Reinforcing. It slows the loop that pays people to rush, which no safety bonus could outweigh.
Near misses on screens: 6 Information. It puts a signal the crews never saw in front of the people who can act on it.
Stop line, keep credit: 5 Rules. It removes the penalty for doing the safe thing, so stopping the line stops costing money.
Goal: injury-free picks: 3 Goals. It changes what the system is for, and every rule and screen gets rearranged to serve it. Expect a fight.
Safe picks are fast: 2 Mindset. It challenges the belief the goal and the bonus came from. Slow to shift, and nothing else holds without it.
An example
A warehouse network's recordable injuries rose for the third year running. The first fix was a number: a bonus for every injury-free shift, raised twice. Nothing changed. The safety lead listed every fix anyone had proposed and placed each one on the twelve points.
The bonus sat at the bottom. Further up sat two things it never touched: the pick-rate bonus paid crews to rush, and near misses reached the site manager a month after the fact, in a summary. Higher still were the rules (a picker who stopped the line lost the shift's pick credit) and the goal itself, picks per hour, posted site against site on every floor screen. She proposed three changes for the year: same-day near-miss reports, stopping the line without losing credit, and a posted goal of injury-free picks per hour. The board approved the first two at once and argued over the third for a quarter.
Why it works
People intervene where the controls are visible, and the visible controls are numbers: a budget, a fine, a target. Numbers matter, but they adjust a system without changing how it works, so the behavior they were meant to fix comes back. Each rung up the ladder changes something the numbers depend on: how fast the system learns of its own results, which loops drive it, who sees what, which rules bind, what it is for, and what the people in it believe.
The ranking also explains the resistance. The higher the point, the more of the system it rearranges and the more people it threatens, so the strongest changes are the slowest and most fought over. Placing every option on one ladder makes the trade visible: an easy fix low down, or a hard one high up.
The thinking behind it
Donella H. Meadows Donella H. Meadows, an environmental and systems scientist and lead author of The Limits to Growth (1972), published her essay Leverage Points through the Sustainability Institute in 1999. She ranked twelve places to intervene, from changing numbers up through rules and goals to changing paradigms, and above that, holding no paradigm as final. Her Thinking in Systems (2008), edited by Diana Wright after her death, explains the stocks, flows, feedback loops, and delays the ranking rests on.
How to run it
Name the outcome that will not move. One measure and its trend: recordable injuries, up three years running.
Map the system behind it. Sketch the stocks, flows, loops, and rules that produce the outcome before listing any fixes.
List every fix on the table. Include the ones already tried; the failures show where the effort has been going.
Place each fix on the twelve points. Ask what it changes: a number, a delay, a loop, who knows what, a rule, the goal, or a belief.
Look at the empty rungs. Points with no fix on them are often where the problem lives.
Favor the highest point you can win. Higher points change more but take longer and draw more resistance, so be honest about the politics.
Pair a high point with the ones under it. A new goal needs the rules and information that let people act on it.
Where it goes wrong
The top of the ladder gets declared, not changed. A leader announces a new mindset or a new goal while the bonus, the rules, and the screens stay as they were, and people follow the incentives. When the slogan changed and the pay plan did not, go back and change the rules underneath.
The ranking gets argued instead of used. Teams debate whether a fix is a six or a five while the outcome keeps sliding. The exact rung matters less than seeing that every fix tried so far sits at the bottom.
High points get planned on a low point's timeline. A goal change scheduled for one quarter will be read as a failure in that quarter. Budget the time and the coalition it needs, and run a lower fix alongside it.
When to use something else
Causal Loop Diagram. When the loops behind the outcome have not been drawn yet, which has to come before placing any fix.
Iceberg Model. When the job is tracing one recurring event down to the structure and beliefs that produce it.
Theory of Constraints. When the system is a line of steps and one bottleneck sets the pace of the whole.