The Blueprint
Define what you want to build.
Before you talk to an AI about building something, you need to know what you want to build.
This sounds obvious. It is not.
Most people start with "I want an app that..." and then wave their hands vaguely. The AI does its best to interpret the hand-waving. The result is something that technically matches what you said but isn't what you meant.
This module is about getting clear. Not perfect. Not complete. Clear enough that an AI can build something you'll recognize.
The One-Sentence Test
Can you describe what you want to build in one sentence? Not a paragraph. Not a pitch deck. One sentence.
If you can\'t, you're not ready to build yet. You\'re still figuring out what you want. That\'s fine. But do the figuring out first.
Good one-sentence descriptions:
A website that lets people RSVP to my workshop and shows a waitlist when spots fill up.
An app that tracks which library books I've lent to friends and reminds me to get them back.
A simple page that converts long YouTube URLs to short ones for sharing.
A tool that takes my rough meeting notes and turns them into formatted action items.
Not-yet-ready descriptions:
Something for my business.
An app like Uber but for [something].
A platform that disrupts [industry].
The next big thing in [category].
A tool that does everything I need.
Your Turn: One Sentence
Write your one-sentence description. Don't overthink it. First draft.
I want to build: _______________________________________________
Now read it back. Does it pass these tests?
- Someone who's never heard of your idea could understand what it does
- It describes ONE thing, not several things
- It\'s specific enough that two different builders wouldn\'t make totally different things
- You could tell if it was "done" or not
If you can\'t pass the tests: That\'s information. You\'re not ready to build yet - you're ready to think. Spend 10 minutes brain-dumping everything you might want this thing to do, then pick the ONE most important part. That's your v1. Everything else is v2.
Who Is This For?
Every app has a user. Who's yours? And I mean specifically - not "everyone" or "people who like productivity."
Who will use this thing?
Be specific. Name a real person if you can: _______________________________________________
What problem does this solve for them?
What's annoying/hard/broken that this fixes? _______________________________________________
How do they solve this problem right now (without your app)?
Spreadsheets? Paper? Memory? Nothing? _______________________________________________
Why this matters: When you're iterating with the AI and something feels "off" but you can\'t explain why, the answer is usually: it doesn\'t match how the actual user would use it. Knowing your user helps you give better feedback. "A librarian wouldn\'t click there" is more useful than "I don't like it."
What Does It Actually Do?
Time to get specific. What are the actions someone can take with this thing?
Think in verbs: search, filter, add, edit, delete, share, export, calculate, display, submit, etc.
List the main actions (start with the most important):
- _______________________________________________
- _______________________________________________
- _______________________________________________
- _______________________________________________
- _______________________________________________
The brutal truth about v1: Your first version should do 1-3 things. That\'s it. If you listed 10 features, circle the top 3. Cut the rest. You\'re not abandoning them - you're saving them for later. The fastest way to build something that does 10 things well is to first build something that does 1 thing well, then add 9 more things.
Circle your TOP 3 actions. Those are your v1.
What Does It Look Like?
You don\'t need to be a designer. But you do need to have some sense of what you're imagining.
When you picture this thing, what do you see?
- A single page (everything visible at once)
- Multiple pages/screens (click to navigate between them)
- A form (user fills in fields and submits)
- A list or table (showing multiple items)
- A dashboard (showing stats or status)
- A tool/calculator (input something, get output)
- Something else: _______________________________________________
Any visual references? (Apps, websites, or tools that have a similar vibe)
"Like Notion but simpler" or "Like that one Google form" - anything helps: _______________________________________________
Any strong visual preferences?
- Dark mode / light mode / either
- Specific colors? _______________________________________________
- Clean and minimal
- Colorful and playful
- Professional and corporate
- I don't care, just make it work
What Data Does It Need?
Apps are basically: collect data, store data, show data, let people do things with data. What data does yours need?
What information does the user put IN?
Names? Dates? Numbers? Text? Files? Selections from a list? _______________________________________________
What information does the user get OUT?
Calculations? Filtered lists? Reports? Confirmations? _______________________________________________
Does anything need to be SAVED between sessions?
- No - it's a one-time calculation or display
- Yes - user accounts and saved data
- Yes - but just locally on their device
- I'm not sure yet
V1 shortcut: If you're not sure about data storage, start with "no" for v1. Build something that works in a single session without saving anything. If people love it and want to save their work, that tells you v2 needs accounts. Don\'t build login systems until you know people want to come back.
Write Your Blueprint
Take everything from the last few pages and combine it into one description. This is what you'll give to Claude.
Use this template or write your own. The goal is: someone reading this could build what you're imagining.
Blueprint Template
I want to build: [one-sentence description]
It's for: [who uses it and why]
The main actions are: [top 3 things users can do]
It should look like: [visual description or reference]
For v1, keep it simple: [any constraints or "don't worry about X yet"]
Now write YOUR blueprint. Full sentences. As if explaining to a smart person who's never heard of your idea.
Read it back: Would someone reading this know what to build? If you showed this to a stranger, would they understand what the end result should look like and do? If yes, you're ready for Module 2. If no, keep refining.
Parking Lot
All the ideas that didn\'t make it into v1. Don\'t delete them - save them here. They're your v2, v3, v4.
Features for later: _______________________________________________
Questions I still have: _______________________________________________
Module 1 Complete
You have a blueprint.
You know what you want to build.
You know who it's for and what it does.
Now it's time to build it.
What\'s next - Module 2: Module 2 is called "Your First Conversation" and it\'s where you actually talk to Claude. You'll open claude.ai, paste your blueprint, and watch something get built. The hard part (figuring out what you want) is done. The fun part is next.
Take a break if you need one. The blueprint isn't going anywhere.