07 - Fair Evaluation Guidelines (Đánh giá công bằng)
Mục tiêu: Đánh giá đóng góp công bằng, không máy móc
Thời lượng: 60 phút Đối tượng: HR, Delivery Manager, Team Lead
🎯 Nguyên tắc vàng: Data + Context + Human Judgment
Formula:
Performance Evaluation =
60% GitHub Data (objective)
+ 30% Tech Lead Review (expertise)
+ 10% Team Feedback (collaboration)
KHÔNG ĐƯỢC:
❌ 100% data (máy móc, thiếu context)
❌ 100% cảm giác (thiên vị, không khách quan)
📊 Framework đánh giá 360°
Step 1: Collect GitHub Data (60%)
Metrics tự động từ GitHub:
✓ Issues completed
✓ Story points delivered
✓ Rework rate
✓ Bug rate
✓ Reviews given
✓ Cycle time
Output: Quantitative score (0-100)
Step 2: Tech Lead Review (30%)
Tech Lead đánh giá:
✓ Code quality (review comments)
✓ Technical complexity handled
✓ Mentorship contribution
✓ Problem-solving ability
✓ Initiative & ownership
Output: Qualitative assessment (1-5 scale)
Step 3: Team Feedback (10%)
Peer review:
✓ Collaboration
✓ Communication
✓ Helpfulness
✓ Team player attitude
Output: Peer rating (1-5 scale)
🧮 Scoring System
Individual Performance Score (IPS)
IPS = (GitHub Score × 0.6) + (Tech Lead Score × 0.3) + (Team Score × 0.1)
Scale: 0-100
90-100: Exceptional
80-89: Exceeds expectations
70-79: Meets expectations
60-69: Needs improvement
< 60: Underperforming
Example: Dev Alice
## Alice - Q2 2024 Performance
### 1. GitHub Data (60%)
- Issues: 32 completed (Avg team: 28) → 114%
- Story points: 165 (Avg: 140) → 118%
- Rework rate: 12% (Target: < 20%) → Excellent
- Bug rate: 8% (Target: < 15%) → Good
- Reviews: 48 (Avg: 30) → 160%
GitHub Score: 92/100
### 2. Tech Lead Review (30%)
- Code quality: 5/5 (Clean, well-tested)
- Complexity: 5/5 (Core auth migration)
- Mentorship: 4/5 (Guided 2 juniors)
- Problem-solving: 5/5 (Solved critical OAuth issue)
- Initiative: 5/5 (Proposed & led refactoring)
Tech Lead Score: 96/100 (4.8/5 × 20)
### 3. Team Feedback (10%)
- Collaboration: 4.5/5
- Communication: 5/5
- Helpfulness: 5/5
- Team player: 5/5
Team Score: 95/100
---
**Final IPS:**
(92 × 0.6) + (96 × 0.3) + (95 × 0.1)
= 55.2 + 28.8 + 9.5
= 93.5/100
**Rating: Exceptional**
**Justification:**
- Consistent high performance
- High-complexity work (core auth)
- Strong collaboration & mentorship
- Zero production bugs
⚖️ Fair Comparison Framework
Rule 1: Compare within SAME role & level
✅ ĐÚNG:
Senior Backend A vs Senior Backend B
❌ SAI:
Frontend Junior vs Backend Senior
(Different stack, different complexity)
Rule 2: Adjust for complexity
Dev A: 50 points (10 simple UI tasks)
Dev B: 40 points (2 complex backend refactors)
❌ SAI: Dev A productive hơn (50 > 40)
✅ ĐÚNG:
Dev A: Average complexity tasks
Dev B: High complexity, high impact
→ Cần normalize by complexity
Complexity Multiplier:
Complexity factor:
- Low (UI changes, simple CRUD): 1.0x
- Medium (business logic, integration): 1.5x
- High (core refactor, architecture): 2.0x
- Critical (security, performance): 2.5x
Adjusted points:
Dev A: 50 × 1.0 = 50
Dev B: 40 × 2.0 = 80
→ Dev B higher impact
Rule 3: Consider tenure & growth
Junior (< 1 year):
- Lower expectations
- Focus on growth rate
- Mentorship needed
Mid (1-3 years):
- Standard expectations
- Independence
Senior (> 3 years):
- High expectations
- Mentorship expected
- Leadership
Example:
Junior Dev Charlie (6 months):
- 20 points/sprint
- Rework rate: 40%
- But: Growth from 10 → 20 points (100% growth!)
→ Rating: Meets expectations (for junior)
→ Focus: Continue learning
Senior Dev David (5 years):
- 35 points/sprint
- Rework rate: 15%
- Mentored 3 juniors
- Led architecture design
→ Rating: Exceeds expectations
🎯 Case Studies: Fair Evaluation
Case 1: Dev with few commits but high impact
Dev Emma:
- Commits: 80 (Low compared to team avg 150)
- Issues: 6 (Low compared to avg 10)
- Story points: 48 (High! Avg 35)
Context:
- Worked on critical payment gateway migration
- High complexity, high risk
- Zero bugs, smooth launch
- Saved company $50k/month
Evaluation:
❌ Naive: Low commits → low productivity
✅ Fair: High impact, critical work → Exceptional
Case 2: Dev with many commits but low impact
Dev Frank:
- Commits: 250 (High! Avg 150)
- Issues: 15 (High! Avg 10)
- Story points: 40 (Avg 35)
But:
- Rework rate: 55% (Poor)
- Bug rate: 30% (Poor)
- Code churn: 4.5 (High churn)
- Many commits fixing own bugs
Evaluation:
❌ Naive: High commits → high productivity
✅ Fair: Low quality, high rework → Needs improvement
Case 3: QA with no code commits
QA Grace:
- Commits: 5 (thêm test cases)
- PRs: 3
- BUT:
- Issues reported: 45 bugs
- Bug quality: 95% valid (not noise)
- Prevented 5 critical production bugs
- Test coverage: 85% → 92%
Evaluation:
❌ Naive: Low commits → low productivity
✅ Fair: High impact on quality → Exceeds expectations
→ KHÔNG thể dùng commit count cho QA!
📝 Evaluation Template
Quarterly Review Template
## Performance Review: [Name] - Q[X] 202X
### Role & Level
- Position: [Senior Backend Engineer]
- Tenure: [2 years]
- Team: [Payment Team]
### Quantitative Metrics (GitHub Data)
#### Productivity
- Issues completed: X (vs team avg Y)
- Story points: X (vs avg Y)
- Cycle time: X days (target < 5)
#### Quality
- Rework rate: X% (target < 20%)
- Bug rate: X% (target < 15%)
- First-time approval: X% (target > 70%)
#### Collaboration
- Reviews given: X (target > 10/sprint)
- Response time: X hours (target < 4h)
**GitHub Score: X/100**
---
### Qualitative Assessment (Tech Lead)
#### Strengths
- [Strength 1 with example]
- [Strength 2 with example]
#### Areas of Excellence
- [What they did exceptionally well]
#### Areas for Growth
- [Constructive feedback 1]
- [Constructive feedback 2]
**Tech Lead Score: X/100**
---
### Team Feedback (Peers)
**Collaboration:** X/5
**Communication:** X/5
**Helpfulness:** X/5
**Highlights from peers:**
- [Positive feedback quote 1]
- [Positive feedback quote 2]
**Team Score: X/100**
---
### Overall Performance Score
**IPS: X/100 (Rating: [Exceptional/Exceeds/Meets/Needs Improvement])**
### Key Achievements
1. [Major achievement 1]
2. [Major achievement 2]
3. [Major achievement 3]
### Development Goals (Next Quarter)
1. [Goal 1]
2. [Goal 2]
3. [Goal 3]
### Support Needed
- [What can manager/team provide]
---
**Manager Signature:** ___________
**Employee Signature:** ___________
**Date:** ___________
⚠️ Common Pitfalls
❌ Pitfall 1: Recency bias
BAD:
Chỉ nhìn tháng cuối (tháng 3) mà quên 2 tháng đầu
GOOD:
Xem trend cả quý:
Month 1: 30 points
Month 2: 35 points
Month 3: 32 points
→ Consistent performance
❌ Pitfall 2: Halo effect
BAD:
Dev A fix 1 critical bug → tất cả đều tốt
GOOD:
Dev A:
- Fixed critical bug (excellent)
- But: High rework rate overall (needs improve)
- And: Low collaboration (needs improve)
→ Mixed performance
❌ Pitfall 3: Comparison bias
BAD:
Dev B ở team A (avg 40 pts/sprint)
Dev C ở team B (avg 25 pts/sprint)
Dev B: 42 points → Good
Dev C: 28 points → BAD (so với Dev B)
GOOD:
Compare within team context:
Dev B: 42 vs team avg 40 → Above avg ✓
Dev C: 28 vs team avg 25 → Above avg ✓
→ Both good relative to their teams
🎯 Action Plan sau khi đánh giá
For High Performers (90-100)
Actions:
✓ Recognize publicly
✓ Reward (bonus, promotion consideration)
✓ Give stretch assignments
✓ Mentorship opportunities
✓ Retain (prevent turnover)
For Solid Performers (70-89)
Actions:
✓ Positive feedback
✓ Identify growth areas
✓ Provide learning resources
✓ Set clear goals for next quarter
For Underperformers (< 70)
Actions:
✓ 1-on-1 discussion
✓ Understand blockers
✓ Create improvement plan (30-60-90 days)
✓ Provide support (training, mentorship)
✓ Monitor progress closely
✓ Escalate if no improvement
✅ Checklist: Fair Evaluation
Before finalizing evaluation:
- [ ] Collected GitHub data for full period
- [ ] Got Tech Lead input
- [ ] Collected peer feedback
- [ ] Considered context (complexity, role, tenure)
- [ ] Compared within same level/role
- [ ] Avoided recency bias
- [ ] Checked for halo/horn effects
- [ ] Prepared specific examples
- [ ] Drafted development goals
- [ ] Scheduled 1-on-1 for feedback delivery
✅ Checklist sau khi đọc xong
- [ ] Hiểu formula: 60% data + 30% tech lead + 10% team
- [ ] Biết scoring system (IPS 0-100)
- [ ] Biết 3 rules: Same role, Complexity, Tenure
- [ ] Hiểu 3 case studies
- [ ] Biết evaluation template
- [ ] Biết pitfalls cần tránh
- [ ] Biết action plan cho từng tier
🚀 Tiếp theo: 09-anti-patterns.md - Lỗi thường gặp khi dùng metrics