skill-110（技能 110）

Skill 110: Organizational Design & Scaling

Quality Grade: 94-95/100 Author: OpenClaw Assistant Last Updated: March 2026 Difficulty: Advanced (requires systems thinking, people dynamics understanding)

Overview

Organizational Design is the practice of structuring teams, roles, and responsibilities to maximize effectiveness, clarity, and alignment. As organizations scale, structure becomes a leverage point—the right structure multiplies productivity; the wrong one creates chaos.

This skill covers:

• Team structures (functional, product, matrix)
• Communication patterns and information flow
• Decision-making frameworks and accountability
• Scaling patterns (when to split teams, create new roles)
• Metrics & culture alignment
• Common failure modes and fixes

Part 1: Team Structure Models

Functional Organization

              CEO
       /      |      \
   Engineering  Product  Sales
       |         |       |
    Backend    Design   Enterprise
    Frontend   Research
    DevOps

Pros:

• Deep expertise (all backend engineers in one team)
• Efficient resource allocation
• Clear career progression

Cons:

• Slow feature delivery (waiting for other functions)
• Silos and "not my problem" thinking
• Product decisions require many meetings

When to use: Early stage (<30 people), specialized domains (research, infra)

Product/Feature Organization

              CEO
       /      |      \
   Payment    Auth    Analytics
   Services   (Squad) (Squad)
   (Squad)
      |         |        |
   Backend   Backend   Backend
   Frontend  Frontend  Frontend
   Product   Product   Product
   DevOps    DevOps    DevOps

Pros:

• Autonomous teams (can ship independently)
• Fast iteration (no waiting for shared functions)
• Clear ownership of outcomes
• Easier scaling (replicate squad structure)

Cons:

• Duplication (each squad has backend, frontend, etc.)
• Knowledge silos (tribal knowledge in squad)
• Requires clear APIs between squads

When to use: Growth stage (30-200 people), feature-driven business

Matrix Organization

                CEO
        /       |       \
   Backend   Frontend   Product
     |         |         |
     +----[Auth Squad]---+
     +----[Payment Squad]
     +----[Analytics Squad]

Pros:

• Leverages functional expertise
• Cross-team knowledge sharing
• Flexibility

Cons:

• Confusing reporting (two bosses)
• Slower decisions (multiple stakeholders)
• "Not my problem" between axes

When to use: Avoid if possible; only for complex orgs that need it

Part 2: Communication & Information Flow

Organizational Bandwidth

Your org can sustain N message types simultaneously:

At 10 people: ~5 types of communication
At 50 people: ~10 types
At 200 people: ~20 types
At 1000 people: ~30 types

Message types:

• Daily standup (what did you ship?)
• Weekly planning (what's coming?)
• Monthly business review (metrics, OKRs)
• Quarterly planning (strategy, roadmap)
• Annual strategy (where are we going?)
• Incident response (when things break)
• Hiring (finding people)
• Onboarding (integrating people)
• Career development (growing people)

Rule of thumb: As you scale, eliminate low-value comms to make room for high-value ones.

Information Architecture

Async-first communication:

Tier 1 (documents):
  - Architecture Decision Records (ADRs)
  - Quarterly plans
  - Postmortems
  - RFCs (Requests for Comments)
  
Tier 2 (recordings):
  - All meetings recorded + transcribed + searchable
  - Weekly planning review (video)
  - Monthly all-hands (video)
  
Tier 3 (synchronous):
  - Only meetings that require real-time discussion
  - Code reviews (async first, pair if stuck)
  - Incident response (synchronous)
  - Onboarding (pairing)

Part 3: Decision-Making & Accountability

Decision Frameworks

RACI:

R = Responsible (does the work)
A = Accountable (yes/no call, has final say)
C = Consulted (asked for input)
I = Informed (told after decision)

Example: "Should we migrate to Kubernetes?"

Migrate to K8s decision:
  A: Infrastructure Lead (yes/no call)
  R: DevOps Engineer (implements)
  C: Product Manager (cost impact), Backend Lead (ops burden)
  I: All engineers (this will affect your workflow)

RAPID (Decisions & Discovery):

R = Recommend (proposes)
A = Agree (must align, has veto)
P = Perform (executes after decision)
I = Input (provides info)
D = Decide (final call)

Escalation Paths

Define clear escalation:

Level 1 (Team Lead): Changes affecting one team
         Decision: 5 days, no review needed (lead decides)

Level 2 (Manager): Cross-team impact
         Decision: 10 days, involves affected teams
         
Level 3 (Director): Strategic direction
         Decision: 2 weeks, involves all stakeholders

Level 4 (C-Suite): Business pivots, org structure
         Decision: 1 month, board review if >$1M impact

Part 4: Scaling Patterns

"Two Pizza Team" Rule

Keep teams small enough to fit two pizzas (6-8 people). When team grows, split.

Splitting Pattern:

Before: Payments Team (10 people)
  - Mobile payments
  - Web payments
  - Payment methods (cards, crypto, etc.)
  - Fraud detection
  - Analytics

After: Two teams
  - Payments Platform (5 people)
    - Core payment APIs
    - Fraud detection
    - Analytics
    
  - Payments Applications (5 people)
    - Mobile client
    - Web client
    - Integration with platform team

Roles That Appear at Scale

At 20 people:

• Engineering manager (first management layer)
• Product manager (dedicated, not CEO wearing hat)
• Designer (shared, not all engineers)

At 50 people:

• Staff/Principal engineer (unblocks others, sets standards)
• Tech lead per team
• Marketing manager
• Operations / People operations

At 200 people:

• Engineering director (manages managers)
• VP Product
• Head of Design
• CFO / Finance
• Head of Recruiting
• Head of People / HR

At 1000+ people:

• Specialized roles (security, compliance, legal, etc.)
• Dedicated roles that were part-time at 200
• Bureaucracy risk increases (culture becomes policy-driven)

Part 5: Common Failure Modes & Fixes

"We grew too fast and lost our culture"

Symptoms:

• Engineers don't know each other anymore
• Decisions slow down
• People feel disconnected
• Onboarding chaos

Fixes:

• Explicitly document culture (values, how we work)
• Hire managers who embody culture, not just technical skills
• Create cohort-based onboarding (same week starts together)
• Monthly all-hands where founders tell story (why we exist)

"We're siloed; teams don't talk"

Symptoms:

• Duplicate work across teams
• Surprises when teams integrate
• Knowledge doesn't spread

Fixes:

• Mandate weekly cross-team syncs (15 min, async-first updates)
• Create shared learning channels (post solutions, blockers)
• Rotate engineers between teams (2-3 month stints)
• Shared on-call (different team covers weekends)

"Decisions are slow; we can't move"

Symptoms:

• Everything needs multiple approvals
• Meetings about meetings
• Employees frustrated

Fixes:

• Document decision authority clearly
• Push decisions down (delegate more)
• Default to "yes, fix it later" instead of "no, risky"
• Create decision SLAs (L1 in 5 days, L2 in 10 days)

"Burnout is killing us; people leaving"

Symptoms:

• Frequent turnover
• "I'm always on call"
• Unpaid overtime expected
• No time for learning

Fixes:

• Real on-call rotation (not just one person)
• 20% time for learning / tech debt
• Explicit time off (use it or lose it)
• Measure and celebrate shipped features, not hours worked
• Career growth conversations (quarterly)

Conclusion

Organizational design directly impacts shipping speed, quality, and happiness. As you scale, structure matters more. The right structure aligns incentives, clarifies ownership, and enables autonomy. The wrong structure creates fiefdoms, slow decisions, and burnout.

Key Takeaway: Your org structure is your strategy made visible. If your strategy and structure don't align, your structure wins.