Developer Feedback Collector

Design structured feedback collection processes that generate actionable, specific, and balanced input for performance reviews and growth conversations.

Purpose

Create feedback systems that:

• Generate specific, example-based feedback
• Identify themes across multiple inputs
• Balance positive and constructive feedback
• Protect psychological safety
• Enable data-driven performance discussions

Core Process

1. Design Feedback Prompts

Structured Question Framework:

## Technical Excellence
- What technical contributions has [Name] made that stand out?
- Where could [Name] improve technically?
- Specific example: [Describe situation]

## Collaboration & Communication
- How effectively does [Name] work with others?
- Example of strong collaboration: [Describe]
- Area for growth in collaboration: [Describe]

## Leadership & Impact
- How does [Name] influence beyond their immediate work?
- Example of leadership: [Describe]
- Growth opportunity in leadership: [Describe]

## Overall
- What should [Name] continue doing?
- What should [Name] start/stop doing?
- One specific development suggestion: [Describe]

Prompt Design Principles:

• Ask for specific examples, not generalizations
• Balance positive and constructive
• Frame constructively (growth opportunities, not failures)
• Use behavioral language (observable actions)
• Avoid leading questions

2. Collect Feedback

Who to Ask:

• Direct manager (required)
• Peers (3-5 engineers at similar level)
• Cross-functional partners (PM, design, etc.)
• Skip-level manager (for senior+)
• Mentees (if applicable)

Collection Methods:

• Written surveys (async, allows reflection)
• 1:1 conversations (deeper context)
• Anonymous (for honest constructive feedback)
• Attributed (for accountability and follow-up)

Best Practice: Mix of attributed and anonymous for balanced input

3. Synthesize Themes

Synthesis Process:

1. Read all feedback inputs
2. Identify recurring themes (3+ mentions = pattern)
3. Group examples by theme
4. Note outlier feedback (1-off mentions)
5. Look for contradictions (investigate causes)

Output Structure:

{
  "employee": "Name",
  "feedback_period": "Q4 2025",
  "respondents": 8,
  "themes": [
    {
      "category": "Technical Excellence",
      "theme": "Strong system design skills",
      "frequency": 6,
      "examples": [
        "Designed scalable caching architecture for API - reduced latency 60%",
        "Led database migration with zero downtime"
      ],
      "type": "strength"
    },
    {
      "category": "Collaboration",
      "theme": "Could improve communication in PRs",
      "frequency": 4,
      "examples": [
        "PR comments sometimes terse, hard to understand reasoning",
        "Could explain design decisions more clearly in code review"
      ],
      "type": "growth_area"
    }
  ],
  "actionable_items": [
    "Continue driving architectural initiatives",
    "Work on PR communication: explain *why* in code review comments"
  ]
}

4. Deliver Feedback

Delivery Framework (Manager → Employee):

## Feedback Summary: [Name] - [Period]

### Key Strengths
[Theme 1 with examples]
[Theme 2 with examples]

### Growth Opportunities
[Theme 1 with examples and specific suggestions]
[Theme 2 with examples and specific suggestions]

### Action Plan
1. [Specific action]: [Why] → [How to measure success]
2. [Specific action]: [Why] → [How to measure success]

### Next Steps
- [Discuss in 1:1 on DATE]
- [Set goals for next period]

Delivery Best Practices:

• Start with strengths (build confidence)
• Be specific (examples, not vague criticisms)
• Focus on behaviors (changeable) not traits (fixed)
• Collaborative tone (partner in growth)
• Forward-looking (action plan, not dwelling on past)

Using Supporting Resources

Templates

• templates/feedback-prompts.json - Question bank by competency
• templates/synthesis-template.md - Theme identification framework

Scripts

• scripts/synthesize-feedback.py - Auto-detect themes from responses
• scripts/anonymize.py - Remove identifying information

---

Progressive Disclosure: Advanced synthesis techniques, difficult feedback scenarios in references/.