Research Workflow
Know what you're building before you start.
Overview
When to Research
Research before:
- Starting a new product
- Adding major features
- Pivoting direction
- Entering new markets
Research is optional for:
- Bug fixes
- Small enhancements
- Technical improvements
Research Outputs
| Output | Purpose |
|---|---|
| Product Vision | What problem we solve |
| User Personas | Who we're building for |
| Feature List | What we'll build |
| Competitor Analysis | What already exists |
| Success Metrics | How we'll measure success |
Step 1: Define the Problem
Question: What problem are we solving?
## Problem Statement
**Who**: [Target user]
**What**: [The problem they face]
**Why**: [Why it matters]
**Current**: [How they handle it now]
### Example
**Who**: Software development teams
**What**: Requirements drift, specs become outdated
**Why**: Causes bugs, rework, and miscommunication
**Current**: Manually update docs, often forgotten
Step 2: Identify Users
Question: Who exactly are we building for?
## User Personas
### Persona 1: Solo Developer
- **Role**: Independent developer
- **Goals**: Build features faster
- **Pain Points**: Context switching, forgetting details
- **Tech Level**: High
### Persona 2: Team Lead
- **Role**: Manages 3-5 developers
- **Goals**: Keep team aligned, track progress
- **Pain Points**: Status meetings, outdated docs
- **Tech Level**: Medium-High
Step 3: Analyze Competition
Question: What solutions exist?
## Competitive Analysis
| Solution | Strengths | Weaknesses | Gap |
|----------|-----------|------------|-----|
| Jira | Enterprise features | Complex, manual | No AI |
| Linear | Beautiful UI | Limited specs | No living docs |
| Notion | Flexible | No dev workflow | Not integrated |
### Opportunity
No existing solution combines:
- AI-assisted development
- Living documentation
- Automatic progress tracking
Step 4: Scope Features
Question: What features matter most?
## Feature Prioritization
### Must Have (MVP)
- [ ] Increment creation
- [ ] Task tracking
- [ ] Basic validation
### Should Have
- [ ] Living docs sync
- [ ] GitHub integration
- [ ] Quality gates
### Nice to Have
- [ ] AI code review
- [ ] Custom workflows
- [ ] Team analytics
### Won't Have (Now)
- [ ] Mobile app
- [ ] Slack integration
Step 5: Define Success
Question: How do we know it's working?
## Success Metrics
### Adoption
- Target: 1,000 active users in 6 months
- Measure: Weekly active increments
### Engagement
- Target: 80% completion rate
- Measure: Increments started vs completed
### Satisfaction
- Target: NPS > 50
- Measure: Quarterly surveys
### Efficiency
- Target: 30% faster feature delivery
- Measure: Time from spec to production
Research Document Template
# Product Research: [Product Name]
## Executive Summary
[1-2 sentence overview]
## Problem Statement
[Who, What, Why, Current state]
## Target Users
[User personas with goals and pain points]
## Competitive Landscape
[Analysis of existing solutions]
## Proposed Solution
[High-level approach]
## Feature Scope
[Prioritized feature list]
## Success Metrics
[How we'll measure success]
## Next Steps
[What happens after research]
Research with AI
Use AI to accelerate research:
# Ask AI to analyze
"Research the task management tool market.
Identify gaps that AI-native development
tools could fill. Focus on developer
workflow automation."
# AI provides
- Market size estimates
- Key competitors
- Underserved segments
- Opportunity areas
Common Research Mistakes
Mistake 1: Skipping Research
❌ "I know what users want"
✅ "Let me validate assumptions with research"
Mistake 2: Analysis Paralysis
❌ Researching for months before starting
✅ Time-boxed research (1-2 weeks for major features)
Mistake 3: Ignoring Competition
❌ "Our idea is unique"
✅ "Here's what exists and why we're different"
When Research Is Complete
Move to design when you can answer:
- Who are we building for?
- What problem are we solving?
- What makes us different?
- What features are essential?
- How will we measure success?