Work Rules

---
name: work-rules
description: Data-driven people operations combining freedom with analytics when building high-trust cultures
---

# Work Rules! - Google's Data-Driven People Operations Framework

**Pattern**: People Operations Framework
**Domain**: Executive Management & Strategy
**Practitioner**: Laszlo Bock (Former SVP of People Operations at Google)
**Source**: Work Rules!: Insights from Inside Google That Will Transform How You Live and Lead (2015)
**Complexity**: Organizational Strategy
**Prerequisites**: HR/People Operations responsibility, data analytics capability
**Estimated ROI**: High - Google scaled from 6,000 to 50,000+ employees using these principles

## Core Principle

Apply data-driven, evidence-based approaches to all people decisions. Test HR assumptions rigorously, give employees freedom and trust, and invest heavily in hiring the right people rather than training the wrong ones.

**Central Insight**: Most HR practices are based on tradition, not evidence. Google's approach: quantify current practices, test new ones with experiments, and optimize based on data.

## When to Use

- **Building or scaling** a people operations function
- **Challenging conventional HR wisdom** with data
- **Improving hiring quality** and reducing training costs
- **Creating high-freedom, high-trust** work environments
- **Implementing transparent compensation** and performance systems
- **Designing employee programs** (benefits, perks, development)

**Warning Signs You Need This**: High training costs (evidence of hiring failures), low employee engagement scores, HR decisions based on "that's how it's always done," lack of metrics on people practices.

## Core Framework: Google's 10 Work Rules

### Rule 1: Give Your Work Meaning
**What**: Connect individual work to company mission and societal impact.
**How**: Articulate mission clearly, show how each role contributes, celebrate impact stories.
**Evidence**: Purpose-driven employees are more engaged and productive.

### Rule 2: Trust Your People
**What**: Give slightly more trust, freedom, and authority than feels comfortable.
**How**: Default to open information sharing, minimize approvals, allow experimentation.
**Marquet's Test**: "If you're not nervous, you haven't given them enough freedom."

### Rule 3: Hire Only People Who Are Better Than You
**What**: Invest 95% of HR dollars in recruiting; 5% in training/development.
**How**: Slow hiring with multiple interviews, use structured interviews, hire for potential not just experience.
**Google Standard**: Hire people who will raise the average quality of the team.

### Rule 4: Don't Confuse Development with Managing Performance
**What**: Separate growth conversations (developmental) from evaluation conversations (performance).
**How**: Schedule development 1-on-1s separately from performance reviews.
**Why**: Employees can't be vulnerable about weaknesses during evaluations.

### Rule 5: Focus on the Two Tails (Top & Bottom Performers)
**What**: Disproportionately invest in your best people; help or move out your worst.
**How**: Give top 5% special projects, learning opportunities, recognition. Bottom 5% get clear feedback and improvement plans.
**Data**: Top performers have 10-100x impact of average performers.

### Rule 6: Be Frugal and Generous
**What**: Cut perks that don't improve performance; invest heavily in what does.
**How**: Test perk effectiveness with data (e.g., free food increases collaboration, fancy chairs don't boost productivity).
**Principle**: Frugal on things, generous on people and mission-critical resources.

### Rule 7: Pay Unfairly
**What**: Large pay differences for large performance differences.
**How**: Top performers earn 5-10x more than average performers in same role.
**Why**: Best people create disproportionate value; pay should reflect reality.

### Rule 8: Nudge (Behavioral Economics for Good)
**What**: Use choice architecture to help employees make better decisions.
**How**: Defaults matter (auto-enroll 401k), social proof works (show what peers do), timing matters (benefits during onboarding).
**Example**: Increased 401k participation from 60% to 95% by changing defaults.

### Rule 9: Manage the Rising Expectations
**What**: Continuously improve programs so expectations rise gradually, not suddenly.
**How**: Make many small improvements rather than few large ones; communicate "beta" status.
**Risk**: Once you give a perk, removing it feels like a loss (loss aversion).

### Rule 10: Enjoy! Then Go Back to #1 and Repeat
**What**: Creating great workplace culture is iterative, not one-time.
**How**: Celebrate progress, gather feedback, experiment with next improvement.

## Implementation Steps

**Phase 1: Establish Data Infrastructure (Months 1-3)**
1. Start measuring: engagement surveys, hiring metrics, performance distributions
2. Create People Analytics function (can start with 1 person)
3. Identify 3 biggest HR pain points to test

**Phase 2: Run First Experiments (Months 4-6)**
4. Pick one conventional practice to challenge (e.g., "do we really need management approval for X?")
5. Design A/B test with control group
6. Measure outcomes quantitatively

**Phase 3: Implement Winners (Months 7-12)**
7. Roll out successful experiments organization-wide
8. Document what you learned; share transparently
9. Build feedback loops to monitor long-term effects

## Key Techniques

**Structured Behavioral Interviews**
- Ask all candidates same questions in same order
- Use past behavior to predict future performance
- Score answers using rubric (1-4 scale on defined criteria)
- Aggregate scores across multiple interviewers
- **Result**: 4x more predictive than unstructured interviews

**Peer Feedback & Calibration**
- Managers nominate ratings; peers provide 360° input
- Calibration meetings force stack-ranking discussions
- Data shows: self-ratings are unreliable (everyone is above average in their own mind)

**Transparency by Default**
- Share company financials, strategy, OKRs with all employees
- Default documents to "world-readable" internally
- **Exception**: Only hide truly sensitive info (M&A, personal data)

## Common Pitfalls

**"We're Not Google"** - These principles work at companies 1/100th Google's size. Start small, test locally.

**Data Without Action** - Measuring is useless if you don't run experiments. Data is for decisions, not dashboards.

**Freedom Without Accountability** - High trust requires high transparency. Freedom fails without visible outcomes.

**Copying Perks Not Principles** - Free lunch won't make you Google. Copy the data-driven methodology, not the specific benefits.

**Ignoring Culture Fit** - Hire for mission alignment. Skills can be taught; values rarely change.

## Edge Cases

- **Regulated Industries**: Some transparency limits (financial services, healthcare). Test maximum safe disclosure.
- **Highly Hierarchical Cultures**: Start with one team as pilot; prove results before scaling.
- **Small Companies (<50 people)**: Can't afford dedicated People Analytics. Use simple surveys (Google Forms) and manual analysis.

## Success Metrics

- **Hiring Quality**: % of new hires rated "exceeds expectations" after 1 year (target: >80%)
- **Retention**: Voluntary attrition of top performers (target: <5% annually)
- **Engagement**: eNPS (employee Net Promoter Score) or Googlegeist scores (target: top quartile for industry)
- **Time to Hire**: Days from req open to offer accepted (balance with quality)
- **Training ROI**: Training spend as % of total people costs (should decrease if hiring well)

## Related Patterns

- **OKRs** (goal-setting framework Google adopted from Intel)
- **Nudge** (Thaler & Sunstein - behavioral economics)
- **DACI Decision Framework** (data-driven decision-making)
- **Radical Candor** (Kim Scott, former Google leader - feedback culture)

---

**🧙🏽‍♂️ Wizard's Note**: The most counter-intuitive insight is Rule 3: spend 95% of people dollars on hiring, not training. Most companies do the reverse. Google's data showed that "a huge training budget is evidence you hired the wrong people." The best training is hiring people who already know how to learn.