Skip to content

Latest commit

 

History

History
170 lines (113 loc) · 4.69 KB

brainstorming.md

File metadata and controls

170 lines (113 loc) · 4.69 KB
title description
The Brainstorming Session
Scoping your MVP before writing any code

The Brainstorming Session

TLDR

Before writing code, have a 30-60 minute conversation with AI to scope your MVP. Describe your full vision, then let AI help you identify what you can realistically build in 2-4 weeks to prove the core concept.

This conversation is the highest-ROI activity in your entire project. It prevents building the wrong thing.

The Conversation

Step 1: Describe everything you want

Don't hold back. Tell AI the full vision:

"I want to build a conference networking app where attendees get AI-powered match recommendations, can schedule meetings, message each other, see analytics on their networking, integrate with LinkedIn and Zoom..."

Step 2: Ask for MVP scoping

"That's my full vision. But I want to validate the concept first. What's the minimum I could build in 2-3 weeks to prove the core idea works?"

Step 3: Let AI analyze

AI with extended thinking will:

  • Identify your core value proposition (what's actually unique)
  • Separate must-haves from nice-to-haves
  • Propose a focused MVP
  • Estimate timeline and cost

Step 4: Iterate

Push back. Ask questions:

  • "Is that really minimal enough?"
  • "What if we cut X?"
  • "Do we need user accounts for MVP?"

Keep narrowing until you have something achievable.

Example Output

After 30 minutes of conversation, you should have:

Core insight:

"The unique value is AI matching. Everything else (messaging, scheduling, analytics) is enhancement. Prove matching works first."

MVP scope:

  • Upload attendee CSV
  • AI generates top 10 matches per person
  • Simple page to view your matches
  • Basic feedback collection
  • No user accounts (magic links)
  • No real-time, no messaging, no integrations

Timeline: 2-3 weeks

Cost: ~$250 in tokens

Success criteria: Do people find their matches valuable?

Red Flags During Scoping

"Everything is essential"

"We need video calls AND messaging AND scheduling or it won't work."

Push back: "Can people use Zoom links and manual scheduling for MVP? Let's prove matching works first."

"It's not that complex"

"It's just calendar integration."

Reality check: Calendar integration = OAuth + multiple providers + timezones + recurring events + error handling = 2 weeks alone. Defer it.

"Users won't understand without X"

"We need an onboarding tutorial or users will be confused."

Counter: Ship without it. If users are confused, that tells you exactly what to explain. Don't guess.

Scope creep during discussion

"Oh, and we should also add..."

Stop. Write it down for v1.0. Don't add it to MVP.

What You Walk Away With

By end of brainstorming, document:

1. MVP Scope (one paragraph)

We're building a conference networking MVP that imports attendee data via CSV, generates AI-powered match recommendations, and displays top 10 matches per attendee with explanations. No user accounts, no messaging, no integrations.

2. Success Criteria (2-3 bullets)

  • 70%+ of test users rate matches as helpful
  • At least one organizer expresses purchase intent
  • Core matching runs in <2 seconds

3. Deferred Features (list)

  • User accounts → v1.0
  • Real-time updates → v1.0
  • Messaging → v1.0
  • Calendar integration → Production
  • Mobile apps → Production

4. Technical Decisions (list)

  • React + Node + PostgreSQL
  • Simple tag matching (not vector DB) for MVP speed
  • Deploy on Vercel free tier

This becomes the foundation for your README and ROADMAP.

Prompts That Work

Opening:

I want to build [full description].

My constraints:
- Timeline: [2-4 weeks for MVP]
- Budget: [$200-500 in tokens]
- Skills: [what you can do]

Help me:
1. Identify the core value proposition
2. Scope an MVP that validates just that
3. List what to defer to later phases
4. Estimate timeline and cost

Narrowing:

That still feels big. What's the absolute minimum 
to test whether [core concept] works?

Technical decisions:

For this MVP scope, what tech stack would you 
recommend and why? Keep it simple.

Common Mistakes

Skipping this conversation entirely Starting with "build me X" instead of "help me scope X." You'll regret it.

Not pushing back on AI's first suggestion AI often proposes something too big initially. Ask "what can we cut?"

Forgetting to document the output The brainstorming insights are valuable. Write them down before they fade.

Treating MVP features as temporary MVP code should still be good code. You're not throwing it away—you're building on it.

Next: Documentation Architecture — Setting up the docs that guide development.