Brainstorm Skill

Procedural knowledge for facilitating structured creative ideation sessions.

When to Use

• Running /brainstorm-board [topic]
• User says "let's brainstorm", "explore options", "what are the possibilities"
• Early-stage problem exploration before research or PRD
• Generating multiple solution approaches

Brainstorming Principles

Diverge Before Converging

1. Quantity first - Generate many ideas without judgment
2. Build on ideas - "Yes, and..." not "No, but..."
3. Wild ideas welcome - Extreme ideas often lead to practical innovations
4. Defer judgment - Evaluation comes later

Tyler's Thinking Style

Reference tyler-context.md - Tyler processes through conversation. Ask probing questions, don't solve immediately.

Process

Step 1: Frame the Problem

Ask clarifying questions:

• What's the core problem we're solving?
• Who has this problem? (Which persona?)
• What does success look like?
• What constraints exist?

Create a problem statement:

For [persona], who [situation/need],
the [feature/solution] is a [category]
that [key benefit].
Unlike [current alternative],
our solution [key differentiator].

Step 2: Generate Ideas (Diverge)

Use multiple ideation techniques:

How Might We (HMW) Reframe problem as opportunity questions:

• "How might we make [problem] easier?"
• "How might we eliminate [friction]?"
• "How might we turn [negative] into [positive]?"

Analogies

• "How does [other industry] solve this?"
• "What would [company/product] do?"
• "If this were [physical/digital/game], how would it work?"

Extremes

• "What if we had unlimited resources?"
• "What if we had to ship in 24 hours?"
• "What if users had zero training?"
• "What if AI could do anything?"

Reversal

• "What would make this problem worse?"
• "What's the opposite approach?"

Step 3: Cluster & Theme

Group ideas by:

• Approach type - Automation vs. augmentation vs. information
• User effort - High-touch vs. low-touch
• Technical complexity - Quick win vs. infrastructure investment
• Persona fit - Which ideas fit which users

Step 4: Rough Evaluation

Score each promising idea:

Criteria	Questions
Impact	How much does this reduce user pain?
Confidence	How sure are we users want this?
Effort	How hard to build/test?
Alignment	Does it fit product vision?

Use simple High/Medium/Low ratings.

Step 5: Select Next Steps

Based on evaluation:

• High confidence, high impact → Move to research/validation
• Low confidence, high impact → Need more evidence first
• Quick wins → Consider for near-term roadmap
• Big bets → Document for future consideration

Output Format

# Brainstorm: [Topic]

**Date:** YYYY-MM-DD
**Participants:** [who was involved]

## Problem Statement
[Framed problem]

## Ideas Generated

### Theme 1: [Name]
| Idea | Description | Impact | Effort |
|------|-------------|--------|--------|
| [Idea A] | [Brief] | H/M/L | H/M/L |
| [Idea B] | [Brief] | H/M/L | H/M/L |

### Theme 2: [Name]
...

## Top Candidates

### 1. [Best Idea]
- **Why promising:** [reason]
- **Key risk:** [concern]
- **Next step:** [action]

### 2. [Second Idea]
...

## Parked Ideas
Ideas worth revisiting later:
- [Idea] - [why parked]

## Questions to Answer
1. [Open question needing research]
2. [Assumption to validate]

## Recommended Next Steps
- [ ] [Specific action 1]
- [ ] [Specific action 2]

Save Location

pm-workspace-docs/research/brainstorms/YYYY-MM-DD-[topic].md

Anti-Patterns

🚩 Converging too early - Let ideas flow before evaluating 🚩 Solution-first - Always start with problem framing 🚩 Single idea - Push for at least 5-10 ideas before narrowing 🚩 Ignoring constraints - Note constraints but don't let them kill creativity 🚩 No next step - Always end with actionable recommendations

Integration with Other Commands

After brainstorming:

• Test assumptions → /hypothesis new [name]
• Start initiative → /new-initiative [name]
• Need user evidence → /research [name]
• Ready for definition → /pm [name]