A new breakdown every week — read the newsletter →
PeopleBusinessesTools
Strategies Mental ModelsDecision Tools Business ModelsFrameworksMoats
Learn Book SummariesReading Lists GuidesQuote CollectionsLearning Paths
Artificial IntelligenceNewsletter AboutContact
Decision Tools

Iceberg Model

Look beneath events to patterns, structures and mental models.

Overview

The Iceberg Model says that visible events are only the tip. Beneath them lie patterns of behaviour over time; beneath those, the systemic structures that produce the patterns; and at the bottom, the mental models — beliefs and assumptions — that hold the structures in place. Fixing events without touching the deeper layers guarantees the event returns.

Its value is leverage. The deeper you intervene, the more durable the change: reacting to an event helps once, but changing the structure or the mental model that generates it can end the whole class of problem.

When to use it

The same problem keeps recurring and event-level fixes never make it stop.

How to use it

Step 1

Events — what happened?

Describe the visible incident you are reacting to.

Step 2

Patterns — what trends?

Ask whether this has happened before. Plot the behaviour over time.

Step 3

Structures — what causes the pattern?

Identify the policies, incentives, and relationships producing the trend.

Step 4

Mental models — what beliefs hold it up?

Surface the assumptions that keep the structure in place.

Step 5

Intervene deep

Choose the deepest layer you can realistically change — that is where leverage lives.

Worked example

A team keeps missing deadlines (event). The pattern: every project slips. The structure: estimates are set by managers who reward optimism. The mental model: ‘admitting uncertainty looks weak.’ Fixing individual deadlines never helps; changing the belief that honest estimates are penalised finally does. The leverage was at the bottom of the iceberg.

Common pitfalls

  • Treating every recurring problem at the event level and wondering why it returns.
  • Stopping at structure without questioning the mental models that sustain it.
  • Assuming the deepest intervention is always feasible — sometimes structure is the realistic lever.

Frequently asked questions

Why do event-level fixes fail?

Because the structure that generated the event is untouched, so it produces the same event again.

Which layer should I intervene at?

The deepest one you can actually change — deeper interventions are more durable but harder.

What are mental models here?

The often-unspoken beliefs and assumptions that make the current structure seem normal or necessary.

Related tools