short answer

Systems thinking is a way to understand problems as a whole instead of only looking at isolated parts. It helps you see how different elements influence each other, find the real cause behind a recurring problem, and choose solutions that improve the system instead of temporarily fixing the symptom.

watch the video

This article is adapted from my video, Systems Thinking | 6 mental models to add to your thinking toolbox. Watch the video if you want the full farm example and a more visual walkthrough of how the six mental models connect.

resources

• Drowning in complexity? Use the Complexity to Clarity frameworks to build mental models to make sense of ambiguity and communicate it clearly to others.

• Want one clear-thinking idea in your inbox each week? Join the Speak with Frameworks newsletter for practical ways to think in frameworks, communicate clearly, and make better decisions.

the 10-second version

the real problem

There are two ways of looking at the world: in parts or as a whole.

Most of us have been trained to think in parts. Take something complex, break it down into individual components, study each component, and then assume you understand the bigger thing.

That works for some problems.

But as Aristotle says, "the whole is more than the sum of its parts." Usually, it's the connection between the different parts that shows you how something actually works.

That is where systems thinking comes in.

Systems thinking helps you understand problems as a whole, identify the cause, and stop only treating the symptoms. The real problem is that many decisions are made with linear thinking in a non-linear world. We assume:

If A happens, then B happens, then C happens

If A happens, then B happens, then C happens

If A happens, then B happens, then C happens

But in many real systems, A affects B, B affects C, and C feeds back into A.

That is why the obvious fix often creates a new problem.

mental model 1: non-linear organization

Linear thinking reduces the world to "if-then."

If A happens, then B happens.
If B happens, then C happens

If A happens, then B happens.
If B happens, then C happens

If A happens, then B happens.
If B happens, then C happens

This is kind of what spreadsheets are built to be like. Everything moves in sequence. One cell leads to another. One step follows another.

But the world is more dynamic than that.

In a non-linear system:

A feeds into B.
B feeds into C.
C feeds back into A

A feeds into B.
B feeds into C.
C feeds back into A

A feeds into B.
B feeds into C.
C feeds back into A

So instead of a straight line, you get a cycle.

In the video, I use the documentary *The Biggest Little Farm* as an example. Molly and John move from L.A. to a farm that used to be used for single-crop berry farming. The soil had been depleted, so they tried to bring the farm back to life with biodiversity.

At first, every solution seemed to create a new problem.

They planted cover crops to bring nutrients and water back into the soil. Then snails infested the cover crop garden.

They raised cows and sheep so manure could fertilize the soil. Then the manure attracted a huge fly overpopulation.

They had a pond with ducks and fish. Then a drought reduced the water, duck droppings created toxic algae, and the fish died.

Each problem looked separate.

But the breakthrough came when they realized the parts weren't working separately. They were working together.

The chickens they bought to lay eggs liked eating the maggots in the manure, which helped contain the fly population. The ducks that were creating problems in the pond also loved eating the snails that were destroying the cover crops. Once the ducks came on land, their droppings became fertilizer instead of poisoning the fish.

Many systems are organically anti-fragile. When you see the interconnectedness, you can often allow the solution to happen instead of forcing one isolated fix.

mental model 2: stock and flow

Once you accept that systems are connected, the next question is:

How do I even know what to map?

Start with stock and flow.

Stock is what exists in the system. These are the things that can be added or subtracted.

On the farm, stock includes:

1. Animals.

2. Humans.

3. Plants.

4. Soil.

5. Water.

6. Money.

Flow is the action that changes the amount of stock.

Selling lemons is a flow. It decreases the number of lemons and increases the amount of money made by the farm.

Composting is a flow. It changes waste into soil nutrients.

Hiring is a flow. It increases the number of people who can do the work.

This simplifies the system. There are only two things to look for:

The stock are the things.
The flow connects the things together

The stock are the things.
The flow connects the things together

The stock are the things.
The flow connects the things together

If you're trying to understand a business problem, this can be surprisingly useful.

For example:

The point isn't to map every possible detail. The point is to stop staring at the whole mess and start naming what exists, what changes, and what connects.

mental model 3: the iceberg model

The iceberg model helps you see four levels of reality:

1. Events.

2. Patterns of behavior.

3. Systems.

4. Mental models.

Most of us only see events.

On the farm, the event is simple:

There are snails on the citrus tree

There are snails on the citrus tree

There are snails on the citrus tree

So the obvious fix is:

Pick off the snails

Pick off the snails

Pick off the snails

That is classic symptom treatment. It's easy to do, but the problem usually comes back.

The second level is patterns of behavior. Here, you add time.

Over an extended period, what keeps happening?

On the farm, the pattern was:

Every time they introduced something new, there was a new problem

Every time they introduced something new, there was a new problem

Every time they introduced something new, there was a new problem

The third level is systems. What structures are creating the pattern?

For the farm, the structure was biodiversity. John and Molly were introducing more life into the farm. Every new plant or animal changed the relationships between prey, predators, nutrients, water, and waste.

The fourth level is mental models. What assumptions, beliefs, or values are shaping the system?

For the farm, the mental model was that biodiversity could create sustainable organic farming, helping humans live in harmony with nature instead of trying to control the environment.

That mental model contained the clue.

The question wasn't "How do we control the snail population?" The better question was:

How does nature deal with snails

How does nature deal with snails

How does nature deal with snails

The answer was already inside the system: ducks.

By making that connection, they didn't need full-time staff hand-picking snails off trees. They had ducks doing the work for free and fertilizing the soil at the same time.

That is the power of moving down the iceberg. You stop reacting to the event and start looking for the structure creating it.

mental model 4: bottleneck

Once you've mapped the system, you may see too many problems.

That is where bottleneck thinking helps.

A bottleneck is the place where things get stuck. It is the weakest point slowing down the whole system.

On the farm, one obvious bottleneck was profit. What was delaying them from making money from the crops and animals they had raised?

There were many problems:

1. They didn't have enough chickens to sell more eggs.

2. The soil wasn't fertile enough to plant more crops.

3. They didn't have enough talent with biodiversity and farming knowledge.

4. Pests were destroying 70% of their crops.

All of these problems mattered.

But the biggest bottleneck was the pests destroying 70% of the crops.

This is important because systems thinking can make you more aware of complexity, but awareness alone doesn't tell you what to fix first.

Bottleneck thinking helps you prioritize.

Ask:

Where is the biggest delay?
What is slowing the whole system down?
If we improved only one thing, what would unlock the most progress

Where is the biggest delay?
What is slowing the whole system down?
If we improved only one thing, what would unlock the most progress

Where is the biggest delay?
What is slowing the whole system down?
If we improved only one thing, what would unlock the most progress

In work, the bottleneck might be unclear decision rights, slow approvals, weak positioning, poor onboarding, a messy handoff, or one unresolved strategic question.

The goal is to find the constraint that affects everything else.

mental model 5: second-order thinking

After you identify the bottleneck, the next question is:

Which solution should I choose

Which solution should I choose

Which solution should I choose

This is where second-order thinking helps.

First-order thinking asks:

If I do this, what 

If I do this, what 

If I do this, what 

Second-order thinking asks:

And then what

And then what

And then what

Let's go back to the farm.

First-order thinking says:

The soil is fallow. If we plant cover crops, that will bring nutrients and water back into the soil

The soil is fallow. If we plant cover crops, that will bring nutrients and water back into the soil

The soil is fallow. If we plant cover crops, that will bring nutrients and water back into the soil

Second-order thinking says:

If the soil gets healthier, what else will happen

If the soil gets healthier, what else will happen

If the soil gets healthier, what else will happen

You might get bugs and critters in the soil. Citrus trees may attract snails. If cover crops attract pests, you need to prepare for that before it becomes a crisis.

You won't know all of this automatically. That is the point.

Most people stop at first-order thinking and say, "I'll deal with whatever happens next."

Second-order thinking gives you direction for what you need to know more about. You can start searching:

What kind of pests do citrus trees attract?
What happens when soil health improves?
What are the common side effects of this intervention

What kind of pests do citrus trees attract?
What happens when soil health improves?
What are the common side effects of this intervention

What kind of pests do citrus trees attract?
What happens when soil health improves?
What are the common side effects of this intervention

Use these three questions:

1. What are the likely outcomes?

2. Out of all the possible outcomes, which one do I think will occur?

3. What is the probability that I'm right?

If you can't answer these questions, it may mean the decision is built on shallow knowledge. That is a signal to find more information or talk to someone who understands the system better.

mental model 6: feedback loop

The final mental model is the feedback loop.

A feedback loop helps you design a system so you have information and data showing whether you're moving closer to your goal.

This brings us full circle.

Systems thinking starts with non-linear organization:

A feeds into B.
B feeds into C.
C feeds back into A

A feeds into B.
B feeds into C.
C feeds back into A

A feeds into B.
B feeds into C.
C feeds back into A

A feedback loop makes that cycle visible.

On the farm, some measurements were obvious. They could track crop output, snail infestations, pest damage, water levels, and profit.

But in business and life, the measurements are often less clear.

If your goal is to clarify your thinking, what should you measure?

Use three steps:

1. Define your goal.

2. Articulate the assumptions you have about how to reach that goal.

3. Choose measurements that match those assumptions.

For example, let's say your goal is:

Clarify my thinking so I can make better decisions

Clarify my thinking so I can make better decisions

Clarify my thinking so I can make better decisions

Your assumptions might be:

Frameworks and mental models will help me think more clearly.
Tracking decisions will help me learn from my choices

Frameworks and mental models will help me think more clearly.
Tracking decisions will help me learn from my choices

Frameworks and mental models will help me think more clearly.
Tracking decisions will help me learn from my choices

Then your measurements could be:

1. How many mental models you know.

2. Which mental models you used in a decision.

3. Whether the decision was good after three or six months.

4. What gap existed between the model you used and the situation.

That creates a feedback loop:

Learn mental models.
Use them in decisions.
Review the decisions.
Identify what worked or what was missing.
Learn better mental models

Learn mental models.
Use them in decisions.
Review the decisions.
Identify what worked or what was missing.
Learn better mental models

Learn mental models.
Use them in decisions.
Review the decisions.
Identify what worked or what was missing.
Learn better mental models

This is how you improve the system instead of only hoping you make better choices next time.

try this

Use this exercise when a problem keeps coming back.

1. Name the event
What happened

1. Name the event
What happened

1. Name the event
What happened

2. Look for the pattern
How often does this happen? When does it happen?
What usually comes before it

2. Look for the pattern
How often does this happen? When does it happen?
What usually comes before it

2. Look for the pattern
How often does this happen? When does it happen?
What usually comes before it

3. Map the system
What structures, incentives, handoffs, or relationships keep producing this pattern

3. Map the system
What structures, incentives, handoffs, or relationships keep producing this pattern

3. Map the system
What structures, incentives, handoffs, or relationships keep producing this pattern

4. Name the mental model
What assumptions are shaping the system

4. Name the mental model
What assumptions are shaping the system

4. Name the mental model
What assumptions are shaping the system

5. Find the bottleneck
Where is the biggest delay, constraint, or weak point

5. Find the bottleneck
Where is the biggest delay, constraint, or weak point

5. Find the bottleneck
Where is the biggest delay, constraint, or weak point

6. Think second order
If we make this change, what will probably happen next

6. Think second order
If we make this change, what will probably happen next

6. Think second order
If we make this change, what will probably happen next

7. Build the feedback loop
What will we measure so we know whether the system is improving

7. Build the feedback loop
What will we measure so we know whether the system is improving

7. Build the feedback loop
What will we measure so we know whether the system is improving

Example:

Event:
The team keeps missing project deadlines

Event:
The team keeps missing project deadlines

Event:
The team keeps missing project deadlines

Pattern:
Deadlines slip whenever requirements change late

Pattern:
Deadlines slip whenever requirements change late

Pattern:
Deadlines slip whenever requirements change late

System:
There is no clear approval checkpoint before work begins

System:
There is no clear approval checkpoint before work begins

System:
There is no clear approval checkpoint before work begins

Mental model:
We assume moving fast means starting immediately

Mental model:
We assume moving fast means starting immediately

Mental model:
We assume moving fast means starting immediately

Bottleneck:
Decision clarity before execution

Bottleneck:
Decision clarity before execution

Bottleneck:
Decision clarity before execution

Second-order question:
If we add an approval checkpoint, will it slow us down or reduce rework

Second-order question:
If we add an approval checkpoint, will it slow us down or reduce rework

Second-order question:
If we add an approval checkpoint, will it slow us down or reduce rework

Feedback loop:
Track how often requirements change after kickoff and how much rework happens

Feedback loop:
Track how often requirements change after kickoff and how much rework happens

Feedback loop:
Track how often requirements change after kickoff and how much rework happens

common mistakes

1. Treating the event as the whole problem. The visible issue is usually only the top layer. Ask what pattern, system, or assumption created it.

2. Mapping everything without choosing a bottleneck. Systems thinking can become overwhelming if you never prioritize. The question is not "What is connected?" It is "Which connection matters most right now?"

3. Assuming your first fix will only have one effect. In a non-linear system, every solution can create side effects. Ask what happens next.

4. Measuring what is easy instead of what matters. A feedback loop only helps if the measurement tells you whether the system is improving.

🧪 why this framework

Systems thinking works because many real problems are dynamic, delayed, and interconnected.

When you only look at one part of the problem, you may solve the visible symptom and accidentally reinforce the deeper structure. The iceberg model slows you down enough to ask whether you're looking at an event, a pattern, a system, or a mental model.

Stock and flow make complex systems easier to map because they separate what exists from what changes it. Bottlenecks help you avoid fixing everything at once. Second-order thinking helps you anticipate consequences before you're forced to react to them. Feedback loops help you learn from the system over time.

The practical benefit is simple: you start asking better questions.

Instead of:

How do we fix this

How do we fix this

How do we fix this

You ask:

What keeps creating this?
Where is the bottleneck?
What happens after the obvious fix?
How will we know if the system is getting better

What keeps creating this?
Where is the bottleneck?
What happens after the obvious fix?
How will we know if the system is getting better

What keeps creating this?
Where is the bottleneck?
What happens after the obvious fix?
How will we know if the system is getting better

That shift is what makes systems thinking useful for decision-making, communication, business strategy, learning, and any problem where the same issue keeps returning.

if you enjoyed this...

You might also like: 10 Mental Models Explained