AgentSkillsCN

high-output-management

以杠杆、量化评估与系统化的团队绩效倍增为核心，构建以结果为导向的管理框架。

SKILL.md
--- frontmatter
name: high-output-management
description: Output-oriented management framework focused on leverage, measurement, and systematic team performance multiplication

High Output Management

Overview

Andy Grove's High Output Management (1983), born from his experience building Intel, reframes management as an output-maximization problem using manufacturing principles. The core insight: a manager's job isn't doing work directly - it's multiplying the output of teams through high-leverage activities. Grove's framework systematically identifies where managers add value (leverage), how to measure what matters (indicators), and how to structure work for maximum collective output. The book introduced OKRs (Objectives and Key Results) and remains the foundational text for Silicon Valley management philosophy.

When to Use

  • •Managing teams where your impact should be multiplicative, not additive
  • •Diagnosing why team output is below potential despite individual competence
  • •Prioritizing manager time among competing demands (meetings, reviews, planning)
  • •Designing processes that scale as teams grow
  • •Implementing systematic goal-setting and performance management
  • •Training new managers on fundamental management responsibilities
  • •Restructuring organizations for higher throughput and faster execution

The Process

Step 1: Define Output, Not Activity

Identify what output your team produces (code shipped, customers served, products sold), not inputs (hours worked, meetings attended). Manager's output = team's output + output of influenced neighboring teams. Measure success by collective results, not personal heroics. Example: Engineering manager's output = features shipped by team, not code they personally write.

Step 2: Identify High-Leverage Activities

Find activities where small manager effort produces large team output change. Three types: (a) Many people affected (all-hands meeting), (b) Activity over time affected (coaching session that improves future work), (c) Critical information provided (unblocking decision). Prioritize ruthlessly for leverage. Example: 1-hour planning session preventing 100 hours of wasted work = 100x leverage.

Step 3: Establish Paired Indicators

Measure what matters with physical, countable metrics. Pair every indicator to prevent gaming (quality + quantity, speed + accuracy). Leading indicators predict future performance, lagging confirm results. Review regularly, react to trends not noise. Example: Code commits (quantity) + bug rate (quality), sales closed (result) + pipeline coverage (leading).

Step 4: Optimize for Limiting Steps

Map your workflow like manufacturing: identify bottlenecks constraining total output. Apply resources to expand the limiting step, not non-bottleneck activities. Continuous improvement focused on throughput. Example: If QA is bottleneck, add QA capacity or automate tests before adding more developers.

Step 5: Apply Managerial Leverage to Meetings

Categorize meetings: (a) Process-oriented (1-on-1s, staff meetings) - regular, predictable, high-leverage; (b) Mission-oriented (ad-hoc project meetings) - specific outcome, time-bound. Treat 1-on-1s as subordinate's meeting for their agenda. Keep staff meetings focused on decision-making, not information sharing. Example: Weekly 1-on-1s (30 min preparation + 60 min meeting = multiplier on direct report's 40-hour week).

Step 6: Implement OKRs (Objectives and Key Results)

Set clear objectives (what we want to achieve) with 3-5 measurable key results (how we know we achieved it). Bottom-up and top-down alignment. Public visibility across organization. Regular check-ins, not annual reviews. Grade honestly (0.7 = success means ambitious targets). Example: Objective: "Dominate enterprise market", KRs: 40% enterprise revenue, NPS 50+, 90% retention.

Step 7: Optimize Task-Relevant Maturity

Match management style to individual's competence and motivation for specific task. Low maturity = structured, directive. Medium = communicative, supportive. High = minimal involvement, monitoring only. Delegate based on maturity, not seniority. Example: Senior engineer unfamiliar with new tech stack needs directive style; junior engineer expert in domain needs hands-off style.

Example Application

Situation: 50-person engineering team shipping slowly despite talented individuals. Each manager "helping" by writing code, creating bottleneck in reviews.

Application:

  • •Step 1: Redefined manager output as "team features shipped" not "personal PRs merged"
  • •Step 2: Identified high-leverage activities: architecture reviews (prevents 3 weeks rework), hiring (multiplies team capacity), 1-on-1 coaching (improves long-term output)
  • •Step 3: Paired indicators: Stories completed + bug rate, Velocity + tech debt ratio
  • •Step 4: Found code review was bottleneck (5-day wait times), added dedicated reviewers, created review SLA
  • •Step 5: Restructured meetings: 30-min 1-on-1s every week, 60-min staff meeting for decisions only, eliminated status meetings for written updates
  • •Step 6: Implemented quarterly OKRs visible to entire team, weekly progress check-ins
  • •Step 7: Assessed each engineer's task-relevant maturity, delegated more to high-maturity ICs, increased structure for low-maturity

Outcome: Feature delivery velocity increased 2.3x in 6 months, manager time in code dropped 80% (reallocated to hiring, planning, coaching), team satisfaction increased 35%.

Anti-Patterns

  • •Measuring inputs (hours worked, lines of code) instead of outputs (value delivered)
  • •Manager as "super individual contributor" (doing work instead of multiplying others' work)
  • •Reacting to every indicator fluctuation instead of identifying trends
  • •Optimizing non-bottleneck steps while ignoring limiting factors
  • •Information-sharing meetings with 20 people (send email instead)
  • •Annual goal-setting with no interim check-ins or adjustments
  • •One-size-fits-all management style regardless of task-relevant maturity
  • •Planning as separate activity from execution (planners should be implementers)

Related

  • •OKRs (Objectives and Key Results) - goal-setting framework introduced in this book
  • •Theory of Constraints - manufacturing principle applied to knowledge work
  • •The Manager's Path (Camille Fournier) - modern application of Grove's principles
  • •Radical Candor - communication framework complementing Grove's 1-on-1 structure
  • •DACI/RACI - decision frameworks for Grove's "neighboring organization influence"
  • •Measure What Matters (John Doerr) - OKR implementation guide from Grove's student
  • •Lean Manufacturing - source of Grove's production principles applied to management