Part I Deep Research

Part I: Deep Research

Purpose

Part I generates comprehensive market and technical research to validate your app idea and inform technical decisions. It uses an adaptive questioning system to create research prompts tailored to your experience level.

How It Works

User Classification System

The workflow begins by identifying your technical background:

• A) Vibe-coder - Great ideas but limited coding experience
• B) Developer - Programming experience and technical knowledge
• C) Somewhere in between - Some coding basics, still learning

Question Flow by User Type

For Vibe-coders (8 Questions)

1. App idea and problem description
2. Target user identification
3. Similar apps and current solutions
4. Unique value proposition
5. Essential features (3 must-haves)
6. Platform preference (phone/web/both)
7. Timeline expectations
8. Budget constraints

For Developers (8 Questions)

1. Research topic and project context
2. Specific research questions (3-5 detailed)
3. Technical decisions to inform
4. Scope boundaries (included/excluded)
5. Depth requirements per area
6. Information source prioritization
7. Technical constraints
8. Business context

For In-Between Users (8 Questions)

1. Project idea and current skills
2. Problem definition and target users
3. Research needs (technical and business)
4. Similar solutions analysis
5. Platform preferences
6. Technical comfort zone
7. Timeline and success metrics
8. Budget for tools and services

AI Model Recommendations

The workflow recommends specific AI models for research:

Best Models for Research (2025)

• Gemini 2.5 Pro - Best for comprehensive research (1M token context window)
• Claude 4.1 Opus - Most accurate technical analysis
• GPT-5 - Good for quick research iterations

Free Access Options

• AI Studio - Free Gemini 2.5 Pro access
• ChatGPT - Limited GPT-5 access
• Claude 4 Sonnet - Limited free access

Generated Research Prompts

Based on user responses, the system generates customized research prompts that include:

For Vibe-coders

• Simple, actionable insights with examples
• 2025 tool recommendations
• Step-by-step implementation guidance
• Cost estimates with free/paid options
• Examples of similar successful projects
• AI agent recommendations for building

For Developers

• Comprehensive technical research
• Architecture patterns and best practices
• Performance benchmarks
• Security considerations
• AI tool integration strategies
• Cost optimization approaches

For In-Between Users

• Balanced practical guidance with technical details
• Implementation complexity analysis
• Time to market comparisons
• Cost-benefit analysis
• Learning resources prioritization

Research Areas Covered

The generated prompts typically address:

1. Market Analysis

   • Competitor landscape
   • User needs validation
   • Market size estimation
   • Pricing strategies

2. Technical Analysis

   • Technology stack recommendations
   • Architecture patterns
   • Development complexity
   • Scaling considerations

3. Implementation Analysis

   • Tool recommendations
   • Development approaches
   • Timeline estimation
   • Resource requirements

4. Cost Analysis

   • Development costs
   • Operating expenses
   • Tool subscriptions
   • Scaling costs

Output

Part I produces a comprehensive research document containing:

• Market validation findings
• Competitor analysis
• Technical recommendations
• Cost estimates
• Implementation roadmap
• Tool recommendations
• Risk assessment

Best Practices

1. Be honest about your technical level - This ensures appropriate question complexity
2. Provide detailed answers - More context leads to better research prompts
3. Use recommended AI models - Gemini 2.5 Pro offers superior research capabilities
4. Save your research - This document informs all subsequent workflow parts
5. Review thoroughly - The research shapes your entire project approach

Time Investment

• Question answering: 5-10 minutes
• AI research generation: 10-20 minutes
• Total time: 15-30 minutes

Next Steps

After completing Part I:

1. Save research findings as research-[AppName].txt
2. Proceed to Part II: PRD Creation
3. Reference research findings when answering PRD questions