How I Decide If an Idea Is a App, a Feature, or Just a Note in Obsidian

Outline

Hook: “I had a great app idea!” I hear this (from myself) 5 times a week. 99% of the time, it’s not an app. It’s a feature. Or a script. Or just a thought that belongs in a note. Knowing the difference saves months of wasted effort.

Core Argument: Most “app ideas” aren’t apps—they’re features, workflows, or notes disguised as apps. A proper decision framework prevents you from building the wrong thing or over-engineering a simple problem. The key: Match the solution complexity to the problem scope.

Key Sections:

The Idea Hierarchy
- Tier 1: Note → Thought, concept, question. No action yet.
- Tier 2: Script/Automation → One-time or periodic task. Set it and forget it.
- Tier 3: Feature → Solves specific problem within existing workflow.
- Tier 4: Tool → Standalone utility, simple interface, focused purpose.
- Tier 5: App/Platform → Multiple workflows, user accounts, data persistence.
- Most ideas: Tier 1-2. Most people try to build: Tier 5.
The Decision Framework: 7 Questions
- Q1: Who else has this problem? (If just you → Script, not app)
- Q2: How often does it happen? (Rare → Script, Frequent → Tool)
- Q3: Does it require state/data over time? (No → Script, Yes → App)
- Q4: Is it part of an existing workflow? (Yes → Feature, No → Tool)
- Q5: Would people pay for it? (No → Note, Maybe → Tool, Yes → App)
- Q6: Can it be solved with existing tools + automation? (Yes → Script)
- Q7: Does it need a UI or can CLI/automation work? (CLI → Script, UI → Tool/App)
- Scoring system: 0-2 points per question
When It’s Just a Note (Stay in Obsidian)
- It’s a concept, not a problem
- No clear user or use case yet
- Might be interesting “someday”
- Example: “What if AI could…” → Note, not project
- Action: Capture in 99 Minds/Obsidian, revisit quarterly
When It’s a Script (Write Code, Not an App)
- Solves your personal problem
- Runs once or on schedule
- No UI needed (CLI is fine)
- Example: Auto-organize downloads folder → Python script + cron
- Action: 50 lines of code, done in 2 hours
When It’s a Feature (Add to Existing Tool)
- Fits naturally into something you’re already building
- Isolated scope, doesn’t require new architecture
- Example: Add “export to PDF” to existing app
- Action: 1-2 day feature addition, not new project
When It’s a Tool (Build Focused Utility)
- Solves one problem really well
- Multiple people have the problem
- Simple UI, no accounts needed
- Example: Image resizer, URL shortener, markdown previewer
- Action: Weekend project, maybe small SaaS
When It’s an App/Platform (Full Build)
- Solves multiple related problems
- Requires user accounts, data storage
- People would pay subscription for it
- Complex enough to have competition
- Example: 99 Minds, project management tool, CRM
- Action: Months of work, proper product strategy
The “Wait One Week” Test
- Write down the idea
- Don’t start building
- Review in 7 days
- Still excited? Move forward. Already forgot? It was a note.
- Prevents impulsive over-building

Examples/Stories:

Idea → Note: “AI for meal planning” → Too vague, just a note
Idea → Script: “Auto-backup my Obsidian vault” → 20-line Python script
Idea → Feature: “Add voice memos to 99 Minds” → Feature, not new app
Idea → Tool: “Simple markdown to PDF converter” → Weekend project
Idea → App: “99 Minds” → Full platform, months of work
False start: Almost built full app for something that needed a script

Takeaways:

Most ideas: Notes or scripts, not apps
Ask 7 questions to properly categorize
Match solution complexity to problem scope
The “wait one week” test prevents over-building
It’s faster to write a script than build an app for a script-sized problem

Cross-Links:

← “The Narrow But Complete Rule” (Series 2-12)
→ “Reverse Real Estate Matchmaking” (Series 2-14)
→ “Idea Cemeteries” (Series 2-17)
← “What I Learned Building 99 Minds” (Series 1-3)
→ “The 99 Minds Principle” (Series 3-26)