7.1 Write a PRD: Turn Your Idea into a Blueprint
Why Write a PRD?
You have a great idea in your head and want to open your editor right away.
Hold on.
Starting to code without a PRD is like driving without a map — you'll eventually end up somewhere, but probably not where you intended, and you'll waste a lot of gas getting there.
A PRD (Product Requirements Document) is your map. It helps you think through three things:
- What exactly are you building (not a vague feeling, but a concrete feature list)
- Who are you building it for (who are your target users, what are their pain points)
- What counts as "enough" (where's the MVP boundary)
"But I'm working alone, not in a team. Do I really need a PRD?"
Yes. Not for others to see — for yourself. Your memory isn't as good as you think. The logic you have crystal clear today will be fuzzy in three days. A short PRD is your "external hard drive."
PRD Template for Small Projects
Enterprise PRDs can run dozens of pages, but for personal projects, one page is enough. Here's our PRD template:
# [Project Name] PRD
## One-Line Description
Explain what this product does in one sentence.
## Problem & Opportunity
What problem are users facing? What's wrong with existing solutions?
## Target Users
Who will use this product? Describe them in one sentence.
## Core Features
List 3-5 core features — don't overdo it.
## User Stories
Use the format "As a [person], I want to [do something], so that I can [get something]."
## MVP Scope
Clearly list: what's in, what's out.
## Tech Choices
Frontend, backend, database, deployment platform.
## Success Criteria
How do you judge this project as "successful"?
You can copy this template and start using it right now. Let's write a real PRD with it.
Hands-On: Writing a PRD for "AI Daily Focus Helper"
Our project is called "AI Daily Focus Helper". Here's the complete PRD:
One-Line Description
A minimalist web app that helps you set 3 priority tasks each day, receive AI encouragement and advice, and track your progress.
Problem & Opportunity
Ever had that feeling? You're busy all day, then at night you look back and realize you didn't do a single important thing.
The problem isn't that you're not working hard enough — it's that you didn't think through "what's most important today" at the start of the day.
There are tons of to-do tools out there (Todoist, TickTick, Notion…), but they have two problems:
- Too complex — too many features, which actually causes anxiety
- No emotional support — just cold, hard task lists with no encouragement or guidance
Let's build a minimalist version: focus on just 3 things per day, plus AI cheering you on.
Target Users
- Knowledge workers aged 25-40 (programmers, designers, operations, freelancers)
- People who have lots to do each day but often get lost in trivial tasks
- Want a simple tool to help them stay focused
- Interested in AI features but don't need anything too complex
Core Features
| Feature | Description | Priority |
|---|---|---|
| Daily Top 3 | Set 3 most important tasks every morning | P0 (must have) |
| Checkbox completion | Check off tasks when done, track progress | P0 |
| AI daily encouragement | Click a button for AI-generated encouragement and tips | P0 |
| History | View past days' tasks and completion status | P1 (nice to have) |
| Streak counter | Show consecutive completion days | P1 |
| Dark mode | Support for nighttime use | P2 (cherry on top) |
User Stories
Story 1: Setting Daily Tasks
As a busy programmer, I want to spend 2 minutes each morning writing down the 3 most important things for today, so that I don't get overwhelmed by trivial tasks.
Story 2: Getting AI Encouragement
As someone who tends to feel anxious, I want to receive encouragement and advice after setting my tasks, so that I'm more motivated to complete them.
Story 3: Tracking Completion
As someone trying to build good habits, I want to see my daily task completion and streak count, so that I can stay motivated.
Story 4: Reviewing the Past
As a reflective person, I want to review my task completion over the past week, so that I can understand my patterns and improve.
MVP Scope
✅ MVP includes:
- A main page where you can enter 3 tasks
- Each task has a checkbox to mark as complete
- A "Get AI Encouragement" button that calls the OpenAI API
- Data saved to Supabase (works without login using localStorage for anonymous users)
- Responsive design, works on mobile too
❌ MVP excludes:
- User login/registration system (anonymous IDs are enough for MVP)
- Multi-user collaboration
- Push notifications/reminders
- Task categories/tags
- Complex data analysis charts
"What you don't build" is just as important as "what you do build." The most common beginner mistake is adding too many features, resulting in none of them being done well.
Tech Choices
| Layer | Technology | Why |
|---|---|---|
| Frontend framework | Next.js 14 (App Router) | Full-stack framework, frontend and backend in one, lots of learning resources |
| Styling | Tailwind CSS | No need to write CSS files, just use utility classes in HTML |
| Database | Supabase (PostgreSQL) | Free tier is enough, comes with API, no need to write backend code |
| AI capabilities | OpenAI API | The most mature AI API, just a few lines of code to call |
| Deployment | Vercel | Next.js's parent company, one-click deployment, free tier |
Why this tech stack?
- All JavaScript/TypeScript — no language switching needed
- All have free tiers — zero cost during development
- All have tons of tutorials — search for answers and someone has already answered
Success Criteria
The success criteria for MVP are simple:
- ✅ You can use it yourself for a full day's workflow (set tasks → get encouragement → check off completions)
- ✅ You can share it with a friend who can understand and use it
- ✅ The entire project is completed in 2 days (entirely feasible with AI-assisted coding)
How to Use AI to Help Write a PRD
Writing a PRD itself is something AI can help with. Here's how:
Step 1: Brainstorm with AI
Open Claude or ChatGPT and type:
I want to build a minimalist to-do app that only lets you set 3 priority tasks per day,
plus an AI encouragement feature. Help me think through:
1. What core problem does this product solve?
2. Who is the target user?
3. What should the MVP include? What should it exclude?
4. What are the potential risks?
AI will give you a great starting point, but remember: the final decision is yours. AI's output is raw material — you decide what to use and what to throw away.
Step 2: Have AI Fill in the Gaps
Share your initial ideas with AI:
Here's my PRD draft:
[paste your content]
Help me check:
1. Are there any important sections I'm missing?
2. Are the user stories complete?
3. Is the MVP scope reasonable? Any unnecessary features?
Step 3: Do the Final Review Yourself
AI might suggest adding many features that "sound nice." This is when you need to be ruthless about cutting. Ask yourself:
"If I had to launch tomorrow, could I finish all these features?"
If the answer is "not sure," move that feature to the "not building" list.
Common PRD Mistakes
Mistake 1: Feature List Too Long
Beginners always want to include every feature they can think of. Remember: The M in MVP stands for Minimum. A 50-feature PRD isn't an MVP — it's a product roadmap.
Mistake 2: User Stories Too Abstract
❌ "As a user, I want a better experience" ✅ "As a user, I want to see a completion animation when I check off a task, so I feel a sense of accomplishment"
Mistake 3: Not Defining "What We Won't Build"
If you don't explicitly state what you're NOT building, you'll gradually start building those features anyway. Boundary awareness is the key to MVP success.
Mistake 4: Following Tech Trends
Don't choose a technology just because it's "hot." Pick what you're familiar with, what has good documentation, and what has an active community. For beginners, "being able to find solutions online" is a hundred times more important than "10% better performance."
Mistake 5: No Success Criteria
If you don't know what "success" looks like, you'll never know when to stop. Set a clear standard for yourself: like "I can finish in 2 days" or "I can share it with 3 friends."
The One-Page PRD Principle
For personal projects, shorter PRDs are better. If your PRD exceeds one A4 page (roughly 500 words of core content), it's probably too complex.
Benefits of a one-page PRD:
- Quick to write — 30 minutes and you're done
- Quick to scan — one glance tells you what to build
- Quick to update — requirements change? Just tweak a few lines
Remember: The PRD isn't homework to turn in — it's for you. If it helps you organize your thoughts, that's enough.
Chapter Summary
| Key Point | Explanation |
|---|---|
| What is a PRD | Product Requirements Document — clarifies "what to build, for whom, to what extent" |
| Template | One-line description + Problem + Users + Features + User stories + MVP scope + Tech + Success criteria |
| What AI can help with | Brainstorming, filling gaps, checking for omissions |
| What you can't rely on AI for | Final decisions, feature trade-offs, priority judgment |
| Core principle | One-page principle — shorter is better; the M in MVP stands for Minimum |
The PRD is done. In the next section, we'll take this blueprint and actually start writing code.