Dunbar's Number

Dunbar’s Number

Cognitive Limits on Social Relationships and Principles for Sizing an Organization

1. Overview: A Principle for Sizing an Organization Around the Cognitive Limits of Social Relationships

    flowchart LR
    A["Organization grows larger<br/>Bureaucratization, broken communication<br/>Trust-based cooperation collapses"] --"Recognize Dunbar's Number<br/>as a cognitive limit"--> B["A layered relationship structure:<br/>5, 15, 50, 150"] --"Optimally design the<br/>organizational unit"--> C["Efficient, trust-based<br/>organizational operation"]

    style A fill:#FFEBEE,stroke:#D32F2F,color:#000
    style B fill:#E3F2FD,stroke:#1976D2,color:#000
    style C fill:#E8F5E9,stroke:#388E3C,color:#000

Definition: A theory derived by British anthropologist Robin Dunbar from research correlating primate neocortex size with group size. It holds that humans can maintain stable social relationships with roughly 150 people at most, and defines relationship capacity by closeness through concentric layers of 5, 15, 50, 150, and 500 — a theory of social cognition.

Characteristics: (The 150-person threshold) Beyond 150 people, it becomes difficult to hold a group together without formal rules, hierarchy, or bureaucracy. (Real-world application) Spotify’s Squad/Tribe model, Amazon’s “two-pizza rule,” and Gore-Tex’s policy of splitting factories at 150 employees are all practical applications of Dunbar’s Number. (Application to remote collaboration) Beyond team size, the cost of maintaining relationships rises further in remote work and asynchronous collaboration, making smaller team design even more important.

2. Core Components of Dunbar’s Number

A. The Layered Structure of Dunbar’s Number

    flowchart TD
    D5["5 people<br/>Close support group<br/>Ask for immediate help in a crisis<br/>Family, closest colleagues"]
    D15["15 people<br/>Sympathy group<br/>Deep trust, shared emotion<br/>Core team, small project"]
    D50["50 people<br/>Band<br/>Capable of cooperation/coordination<br/>Department, squad cluster"]
    D150["150 people<br/>Dunbar's Number<br/>Upper bound of a stable social group<br/>Can operate without formal rules"]
    D500["500 people<br/>Megaband<br/>Face recognition,<br/>weak ties"]
    D1500["1,500 people<br/>Tribe<br/>Name-recognition level,<br/>shared sense of belonging only"]

    D5 --> D15 --> D50 --> D150 --> D500 --> D1500

    style D5   fill:#1E3A5F,stroke:#1E3A5F,color:#fff
    style D15  fill:#E3F2FD,stroke:#1976D2,color:#000
    style D50  fill:#F3E5F5,stroke:#7B1FA2,color:#000
    style D150 fill:#FFF3E0,stroke:#F57C00,color:#000
    style D500 fill:#f5f5f5,stroke:#ccc,color:#555
    style D1500 fill:#f5f5f5,stroke:#ccc,color:#555

Relationship Characteristics by Layer

Layer	Size	Relationship Character	Communication Style	IT Org Application
Close support group	~5	Unconditional trust, immediate support	Frequent conversation, immediate contact	Pair-programming partner, core tech lead
Sympathy group	~15	Deep trust, candid feedback	Daily standup, weekly 1:1	Scrum team, small feature team
Band	~50	Close cooperation, role coordination	Biweekly team meeting, Slack channel	Squad cluster, small department
Dunbar’s Number	~150	Limit of stable trust relationships	Monthly town hall, org newsletter	Small startup, business unit
Megaband	~500	Faces known, but no deep relationship	Quarterly company-wide meeting	Mid-size company business division

B. Application to Organizational Design and Team-Size Optimization

    flowchart LR
    subgraph R1["　"]
        direction LR
        O1["Two-Pizza Rule<br/>Amazon's Jeff Bezos<br/>Team size fed by<br/>two pizzas = 6-10 people"]
        O2["Squad/Tribe model<br/>Spotify model<br/>Squad: 6-12 people<br/>Tribe: 40-150 people"]
    end
    subgraph R2["　"]
        direction LR
        O3["Split factories at 150<br/>Gore-Tex (W.L. Gore)<br/>Build a new factory once<br/>headcount exceeds 150"]
        O4["Microteam structure<br/>Netflix/Basecamp<br/>Fast decisions through<br/>small, autonomous teams"]
    end

    style O1 fill:#E3F2FD,stroke:#1976D2,color:#000
    style O2 fill:#F3E5F5,stroke:#7B1FA2,color:#000
    style O3 fill:#FFF3E0,stroke:#F57C00,color:#000
    style O4 fill:#E8F5E9,stroke:#388E3C,color:#000
    style R1 fill:none,stroke:none
    style R2 fill:none,stroke:none

Communication Channels and Recommended Size by Team Size

Team Size	Number of Channels	Management Complexity	Recommended Structure
3-5	3-10	Very low	Pair or small core team
6-10	15-45	Low	The sweet spot for the Two-Pizza Rule
10-15	45-105	Medium	Maximum size for a Scrum team
15-50	105-1,225	High	Team splitting/sub-team structure needed
Beyond 150	11,175+	Very high	Formal hierarchy/process/rules essential

The 150-Person Rule — When to Split an Organization

    flowchart LR
    G1["Team/department size<br/>under 150"] --"Trust-based,<br/>informal cooperation possible"--> G2["Minimal formal rules<br/>Autonomous operation possible"]
    G3["Team/department size<br/>over 150"] --"Exceeds the cognitive limit<br/>for maintaining relationships"--> G4["Bureaucratization needed,<br/>or split the organization"]

    style G1 fill:#E8F5E9,stroke:#388E3C,color:#000
    style G2 fill:#E8F5E9,stroke:#388E3C,color:#000
    style G3 fill:#FFEBEE,stroke:#D32F2F,color:#000
    style G4 fill:#FFEBEE,stroke:#D32F2F,color:#000

3. Expected Benefits and Practical Application of Dunbar’s Number

Category	Key Expected Benefit	Practical Application
Team design	Improved communication efficiency from sizing teams around cognitive limits	Apply a 6-10 person guideline for agile teams; consider splitting beyond 15
Organizational scaling	Plan a split at the 150-person threshold to preserve culture	Prepare a re-organization roadmap around 150 people as a startup grows
Remote team management	Smaller teams to account for higher relationship-maintenance cost remotely	Structure remote teams as 5-8 people to sustain deep relationships
Microservice team structure	Design microservice teams around the Two-Pizza Rule	Align service boundaries (DDD bounded contexts) and team size with Dunbar’s Number

Tuckman Stages Psychological Safety