spec-driven-brainstorming

规范驱动的头脑风暴技能

产品探索、功能构思和规范驱动头脑风暴技巧方面的专家。通过结构化引导方法，帮助团队将模糊想法转化为具体、定义明确的规范。

核心引导技巧

1. 故事地图（用户故事地图）

目的：可视化用户旅程，并找出每个步骤中能够创造价值的功能。

流程：

Step 1: Define User Activities (horizontal backbone)
┌──────────────┬──────────────┬──────────────┬──────────────┐
│ Discover     │ Browse       │ Purchase     │ Receive      │
│ Products     │ & Compare    │ & Checkout   │ & Review     │
└──────────────┴──────────────┴──────────────┴──────────────┘

Step 2: Break down into User Tasks (vertical slices)
Discover Products:
├─ Search by keyword
├─ Filter by category
├─ View trending products
└─ Get personalized recommendations

Browse & Compare:
├─ View product details
├─ Read reviews
├─ Compare products side-by-side
└─ Save to wishlist

Purchase & Checkout:
├─ Add to cart
├─ Apply discount code
├─ Select shipping method
└─ Enter payment info

Step 3: Prioritize by Walking Skeleton (MVP = top row)
┌────────────────────────────────────────────────────────┐
│ MVP (Release 1): Walking Skeleton                      │
├────────────────────────────────────────────────────────┤
│ Search → View Details → Add to Cart → Checkout        │
└────────────────────────────────────────────────────────┘

┌────────────────────────────────────────────────────────┐
│ Release 2: Enhanced Discovery                          │
├────────────────────────────────────────────────────────┤
│ Filters, Trending, Recommendations, Reviews            │
└────────────────────────────────────────────────────────┘

┌────────────────────────────────────────────────────────┐
│ Release 3: Advanced Features                           │
├────────────────────────────────────────────────────────┤
│ Wishlist, Compare, Discount Codes, Saved Payments     │
└────────────────────────────────────────────────────────┘

输出：与用户旅程相匹配的优先级待办事项列表。

2. 事件风暴

目的：通过协作建模，发现领域事件和业务流程。

流程：

## Event Storming Workflow

### Step 1: Identify Domain Events (orange sticky notes)
- OrderPlaced
- PaymentProcessed
- OrderShipped
- OrderDelivered
- OrderCancelled

### Step 2: Identify Commands (blue sticky notes)
- PlaceOrder
- ProcessPayment
- ShipOrder
- CancelOrder

### Step 3: Identify Aggregates (yellow sticky notes)
- Order (handles PlaceOrder, CancelOrder)
- Payment (handles ProcessPayment)
- Shipment (handles ShipOrder)

### Step 4: Identify External Systems (pink sticky notes)
- PaymentGateway (Stripe)
- ShippingProvider (FedEx API)
- InventorySystem

### Step 5: Identify Policies (purple sticky notes)
- WHEN OrderPlaced THEN ProcessPayment
- WHEN PaymentProcessed THEN ReserveInventory
- WHEN InventoryReserved THEN ShipOrder
- WHEN OrderCancelled AND PaymentProcessed THEN RefundPayment

输出：业务流程和限界上下文的可视化地图。

3. 影响地图

目的：通过用户影响，将业务目标与功能关联起来。

GOAL: Increase revenue by 20% in Q2

WHY? (Impact)
├─ Increase conversion rate (5% → 8%)
│  ├─ WHO? (Actors)
│  │  ├─ New visitors
│  │  └─ Returning customers
│  ├─ HOW? (Features)
│  │  ├─ Simplify checkout (1-click purchase)
│  │  ├─ Add product recommendations
│  │  └─ Offer guest checkout
│  └─ WHAT? (Deliverables)
│     ├─ US-001: 1-click checkout for logged-in users
│     ├─ US-002: ML-based product recommendations
│     └─ US-003: Guest checkout flow
│
├─ Increase average order value ($50 → $65)
│  ├─ WHO? (Actors)
│  │  └─ Existing customers
│  ├─ HOW? (Features)
│  │  ├─ Bundle discounts (buy 3, get 10% off)
│  │  ├─ Free shipping threshold ($75+)
│  │  └─ Upsell related products
│  └─ WHAT? (Deliverables)
│     ├─ US-004: Bundle discount engine
│     ├─ US-005: Dynamic shipping calculator
│     └─ US-006: Related product suggestions
│
└─ Reduce cart abandonment (40% → 25%)
   ├─ WHO? (Actors)
   │  └─ Users with items in cart
   ├─ HOW? (Features)
   │  ├─ Cart abandonment emails
   │  ├─ Save cart across devices
   │  └─ Show trust signals (reviews, secure badges)
   └─ WHAT? (Deliverables)
      ├─ US-007: Automated cart recovery emails
      ├─ US-008: Persistent cart sync
      └─ US-009: Trust badge UI components

输出：与业务成果直接关联的功能。

优先级框架

1. MoSCoW 方法

定义：将功能分为必须有、应该有、可以有和不会有。

## Feature Prioritization: E-commerce Platform MVP

### MUST Have (Critical for Launch)
- [ ] User registration & login
- [ ] Product catalog with search
- [ ] Shopping cart
- [ ] Checkout with payment processing
- [ ] Order confirmation email

**Rationale**: Core transactional flow, no sales without these.

### SHOULD Have (Important but not critical)
- [ ] Product reviews and ratings
- [ ] Wishlist/Save for Later
- [ ] Order history
- [ ] Basic analytics dashboard (admin)

**Rationale**: Enhance UX and trust, but MVP can ship without.

### COULD Have (Nice to have if time allows)
- [ ] Product recommendations
- [ ] Social login (Google, Facebook)
- [ ] Advanced filtering (price range, brand)
- [ ] Guest checkout

**Rationale**: Competitive features, but not required for MVP.

### WON'T Have (Explicitly deferred)
- [ ] Mobile app (web-first)
- [ ] Multi-currency support
- [ ] Subscription billing
- [ ] Loyalty program

**Rationale**: Future roadmap items, not needed for initial market validation.

适用场景：定义 MVP 范围、设定时间限制的发布。

2. RICE 评分（覆盖范围、影响力、信心、投入）

公式：RICE Score = (Reach × Impact × Confidence) / Effort

## RICE Scoring Example

### Feature A: 1-Click Checkout
- **Reach**: 5000 users/month will use this
- **Impact**: High (3/3) - significantly reduces friction
- **Confidence**: 80% (have data from competitor analysis)
- **Effort**: 4 person-weeks

**RICE Score** = (5000 × 3 × 0.8) / 4 = **3000**

### Feature B: Product Recommendations
- **Reach**: 8000 users/month will see recommendations
- **Impact**: Medium (2/3) - incremental revenue lift
- **Confidence**: 50% (no A/B test data yet)
- **Effort**: 8 person-weeks

**RICE Score** = (8000 × 2 × 0.5) / 8 = **1000**

### Feature C: Guest Checkout
- **Reach**: 2000 users/month (30% of visitors)
- **Impact**: High (3/3) - reduces abandonment significantly
- **Confidence**: 90% (industry benchmarks strong)
- **Effort**: 2 person-weeks

**RICE Score** = (2000 × 3 × 0.9) / 2 = **2700**

### Priority Order
1. **1-Click Checkout** (RICE: 3000)
2. **Guest Checkout** (RICE: 2700)
3. **Product Recommendations** (RICE: 1000)

适用场景：数据驱动的优先级排序、路线图规划。

3. Kano 模型

类别：

• 基本需求（必备型）：缺少会导致不满，具备则不会带来惊喜
• 性能需求（线性型）：越多越好（满意度线性提升）
• 兴奋需求（魅力型）：缺少不会造成影响，具备则令人惊喜

## Kano Analysis: Email Client

### Basic Needs (Hygiene Factors)
- Send and receive email (expected, must work flawlessly)
- Attachment support (expected)
- Spam filtering (expected)

**Action**: Must implement, but won't differentiate product.

### Performance Needs (Satisfiers)
- Search speed (faster = better satisfaction)
- Storage quota (more = better satisfaction)
- Mobile app performance

**Action**: Invest proportionally based on competitive benchmarks.

### Excitement Needs (Delighters)
- AI-powered email summarization (unexpected, delights users)
- Smart reply suggestions
- Scheduled send with timezone awareness
- Undo send (5-second window)

**Action**: Focus on 1-2 delighters for differentiation.

### Indifferent Features (Low Priority)
- Custom email signatures (users don't care much)
- Theme customization (low impact)

**Action**: Deprioritize or skip.

### Reverse Features (Causes Dissatisfaction)
- Intrusive ads in inbox (annoys users)
- Forced social features (users resist)

**Action**: Avoid completely.

适用场景：了解客户满意度的驱动因素、制定差异化策略。

精益创业验证

1. 构建-测量-学习循环

## Hypothesis Testing: Feature X

### BUILD
**Hypothesis**: Adding product recommendations will increase average order value by 15%.

**Minimum Viable Test**:
- Implement simple "Customers also bought" section
- Show on 50% of product pages (A/B test)
- Track: clicks, add-to-cart rate, order value

**Effort**: 1 week (backend + frontend)

### MEASURE
**Metrics to Track**:
- Click-through rate on recommendations
- Add-to-cart conversion from recommendations
- Average order value (treatment vs control)
- Revenue per visitor

**Success Criteria**:
- CTR > 5%
- AOV increase > 10%
- Statistical significance (p < 0.05)

**Data Collection Period**: 2 weeks (minimum 10,000 visitors)

### LEARN
**Scenario A: Hypothesis Validated**
- AOV increased 18% (exceeded target!)
- CTR on recommendations: 12%
- **Action**: Roll out to 100%, invest in ML-based recommendations

**Scenario B: Hypothesis Rejected**
- AOV increased 2% (below target)
- CTR on recommendations: 1% (low engagement)
- **Action**: Pivot - test alternative hypothesis (e.g., bundle discounts)

**Scenario C: Mixed Results**
- AOV increased 12% (close to target)
- High CTR but low conversion
- **Action**: Iterate - improve recommendation quality (ML model)

2. MVP 定义画布

## MVP Canvas: Task Management SaaS

### Target Users
- Solo freelancers and small teams (2-5 people)
- Knowledge workers (designers, developers, writers)
- Currently using: Spreadsheets, Trello, Notion

### Problem Being Solved
- Task prioritization is manual and time-consuming
- No visibility into blockers and dependencies
- Team collaboration requires constant status updates

### Unique Value Proposition
Auto-prioritized task list using AI + team workload balancing.

### MVP Features (Walking Skeleton)
**Core Flow**: Create task → AI prioritizes → Assign → Complete

**Must-Have Features**:
- [ ] Task creation (title, description, due date)
- [ ] AI prioritization (urgency + importance algorithm)
- [ ] Task assignment to team members
- [ ] Task status updates (To Do, In Progress, Done)
- [ ] Team dashboard (workload overview)

**NOT in MVP**:
- ❌ Time tracking
- ❌ Custom workflows
- ❌ Integrations (Slack, GitHub)
- ❌ Mobile app
- ❌ Advanced reporting

### Success Metrics
- **Activation**: 70% of signups create 3+ tasks in first week
- **Retention**: 40% weekly active users (WAU) after 4 weeks
- **Engagement**: Average 5 tasks completed/week per user

### Risks & Assumptions
- **Assumption**: Users trust AI prioritization
  - **Test**: Survey 50 users after 2 weeks, ask "Do you trust the priority scores?"
- **Risk**: AI prioritization is inaccurate
  - **Mitigation**: Manual override, feedback loop to improve model
- **Assumption**: Teams of 2-5 are willing to pay $10/user/month
  - **Test**: Offer paid tier after 2-week trial, track conversion rate

头脑风暴技巧

1. 疯狂 8 分钟（快速构思）

流程：8 分钟绘制 8 张草图（每个想法 1 分钟）。

## Crazy 8s Session: Improve Checkout Flow

### Ideas Generated (8 minutes)
1. **1-Click Purchase** - Saved payment + address, single button
2. **Progressive Disclosure** - Multi-step wizard (cart → shipping → payment)
3. **Guest Checkout** - No account required, email-only
4. **Cart Abandonment Recovery** - Email + discount code
5. **Payment Link Sharing** - Send checkout link to someone else (gift)
6. **Buy Now Pay Later** - Installment payments (Klarna integration)
7. **Voice Checkout** - "Alexa, complete my order"
8. **AR Try-On** - Virtual fitting room before checkout

### Voting (Dot Voting)
- 1-Click Purchase: ●●●●● (5 votes)
- Guest Checkout: ●●●● (4 votes)
- BNPL Integration: ●●● (3 votes)
- Progressive Disclosure: ●● (2 votes)

### Top 3 for Deeper Exploration
1. 1-Click Purchase (quick win, high impact)
2. Guest Checkout (reduce friction)
3. BNPL Integration (competitive parity)

2. 六顶思考帽（德·波诺）

目的：从不同视角探索想法。

## Six Hats Analysis: Feature X (AI-Powered Email Summarization)

### White Hat (Facts & Data)
- Average email length: 200 words
- Users spend 3 minutes reading complex emails
- 40% of emails are > 500 words
- Competitor Y launched similar feature (20% adoption)

### Red Hat (Emotions & Intuition)
- "This feels like a gimmick, I don't trust AI to summarize important emails"
- "Love this! Saves time on long threads"
- "Worried about missing critical details in summary"

### Yellow Hat (Optimism & Benefits)
- Saves 2 minutes per long email → 20 min/day for heavy users
- Reduces cognitive load, improves focus
- Differentiator from competitors (if done well)
- Could upsell as premium feature

### Black Hat (Risks & Caution)
- AI hallucination risk (incorrect summaries)
- Privacy concerns (email content processed by AI)
- High development cost (NLP model training)
- May annoy users who prefer full context

### Green Hat (Creativity & Alternatives)
- Alternative 1: Highlight key sentences (instead of summary)
- Alternative 2: TL;DR generated by sender (not AI)
- Alternative 3: Voice-to-summary (read email aloud, generate summary)

### Blue Hat (Process & Conclusion)
**Decision**: Proceed with MVP (limited rollout)
- Build: Highlight key sentences (lower risk than full summary)
- Test: 10% of users, measure engagement + feedback
- Iterate: If successful, invest in full AI summarization

3. 我们如何才能（HMW）问题

目的：将问题重新构想为机遇。

## Problem Statement
Users abandon checkout because the form is too long (12 fields).

### HMW Questions
- **HMW reduce the number of required fields?**
  - Idea: Use address autocomplete (Google Places API)
  - Idea: Prefill from previous orders
- **HMW make the form feel shorter?**
  - Idea: Multi-step wizard (psychological chunking)
  - Idea: Progress bar showing "80% complete"
- **HMW eliminate the form entirely?**
  - Idea: 1-click checkout for returning users
  - Idea: Voice input for address/payment
- **HMW make filling the form more enjoyable?**
  - Idea: Gamify with rewards (10 points per field completed)
  - Idea: Show real-time savings ("You've saved $15 so far!")
- **HMW help users trust the checkout process?**
  - Idea: Show trust badges (SSL, money-back guarantee)
  - Idea: Live chat support during checkout

功能拆解模板

Epic → 功能 → 用户故事

## Epic: User Onboarding Experience

### Feature 1: Account Creation
**User Story US-001**: Email/Password Registration
- **As a** new user
- **I want to** create an account with email/password
- **So that** I can access personalized features

**Acceptance Criteria**:
- Email validation (RFC 5322 format)
- Password complexity (8+ chars, 1 uppercase, 1 number, 1 special)
- Duplicate email detection
- Verification email sent within 5 minutes

**User Story US-002**: Social Login (Google, GitHub)
- **As a** new user
- **I want to** sign up with my Google/GitHub account
- **So that** I don't have to remember another password

**Acceptance Criteria**:
- OAuth 2.0 integration
- Consent screen shown
- Email auto-verified for social logins

### Feature 2: Profile Setup
**User Story US-003**: Basic Profile Information
- **As a** new user
- **I want to** set my display name and avatar
- **So that** other users can recognize me

**User Story US-004**: Preferences Configuration
- **As a** new user
- **I want to** configure notification preferences
- **So that** I only receive relevant updates

### Feature 3: Guided Tour
**User Story US-005**: Interactive Product Tour
- **As a** first-time user
- **I want** a guided tour of key features
- **So that** I understand how to use the product

**User Story US-006**: Sample Data Pre-population
- **As a** new user
- **I want** sample data to explore
- **So that** I can try features without manual setup

协作式研讨会形式

1. 远程头脑风暴（Miro/FigJam）

议程（90 分钟）：

00:00 - 00:10  Introduction & Problem Statement
00:10 - 00:25  Individual Ideation (silent brainstorming)
00:25 - 00:45  Group Sharing (2 min per person)
00:45 - 01:00  Affinity Grouping (cluster similar ideas)
01:00 - 01:15  Dot Voting (3 votes per person)
01:15 - 01:30  Discussion & Action Items

工具：

• 带模板的 Miro 画板
• 用于限定时长的计时器
• 匿名投票

2. 设计冲刺（5 天形式）

Day 1: Map (Understand the problem)
- User journey mapping
- Identify pain points
- Set sprint goal

Day 2: Sketch (Diverge - generate ideas)
- Crazy 8s
- Solution sketches
- Silent critique

Day 3: Decide (Converge - choose solution)
- Dot voting
- Storyboard creation
- Prototype plan

Day 4: Prototype (Build realistic facade)
- High-fidelity mockup
- Interactive prototype (Figma)
- Test script preparation

Day 5: Test (Validate with users)
- 5 user interviews
- Record findings
- Decide: build, iterate, or pivot

输出模板

头脑风暴会议总结

# Brainstorming Session: [Topic]

**Date**: 2024-01-15
**Participants**: Alice (PM), Bob (Eng), Carol (Design)
**Facilitator**: Alice

## Problem Statement
Users are abandoning checkout at 40% rate (industry avg: 25%).

## Ideas Generated (22 total)

### High Priority (Top 5 by voting)
1. **1-Click Checkout** (8 votes)
   - Rationale: Removes friction for returning users
   - Effort: 2 weeks
   - Impact: Est. 10% reduction in abandonment

2. **Guest Checkout** (7 votes)
   - Rationale: 30% of users don't want accounts
   - Effort: 1 week
   - Impact: Est. 8% reduction in abandonment

3. **Progress Indicator** (6 votes)
   - Rationale: Reduces anxiety about form length
   - Effort: 2 days
   - Impact: Est. 3% reduction in abandonment

4. **Autofill Address** (5 votes)
   - Rationale: Saves time, reduces errors
   - Effort: 1 week (Google Places API)
   - Impact: Est. 5% reduction in abandonment

5. **Save Cart for Later** (4 votes)
   - Rationale: Users can return without starting over
   - Effort: 3 days
   - Impact: Est. 4% recovery of abandoned carts

### Medium Priority (Parking Lot)
- Buy Now Pay Later integration
- Live chat support during checkout
- Trust badges (SSL, money-back guarantee)

### Deferred (Low ROI or High Risk)
- Voice checkout (too experimental)
- AR try-on (out of scope)

## Action Items
- [ ] Alice: Create specs for Top 3 (1-Click, Guest, Progress)
- [ ] Bob: Technical feasibility assessment (3 days)
- [ ] Carol: Mockups for guest checkout flow (5 days)
- [ ] Team: Review specs on Friday standup

## Next Session
- Date: 2024-01-22
- Topic: Refine top 3 ideas into user stories

最佳实践

1. 为所有环节设定时限

• 构思：最多 10-15 分钟
• 讨论：每个想法 5 分钟
• 投票：2 分钟

2. 先发散，再收敛

• 先追求数量（不做批评）
• 稍后评估质量（采用结构化投票）

3. 采用可视化方式

• 草图 > 文本
• 白板 > 文档
• 原型 > 规范

4. 纳入多元视角

• 工程（可行性）
• 设计（易用性）
• 产品（业务价值）
• 支持（用户痛点）

5. 记录决策

• 为什么选择 X 而不是 Y？
• 我们做出了哪些假设？
• 我们会衡量什么？

资源

• 用户故事地图 - Jeff Patton
• 影响地图 - Gojko Adzic
• 设计冲刺 - Google Ventures
• Kano 模型分析

激活关键词

可以向我询问：

• “如何开展头脑风暴会议”
• “用于产品探索的故事地图”
• “优先级框架（MoSCoW、RICE、Kano）”
• “如何将 Epic 拆解为用户故事”
• “精益创业验证技巧”
• “MVP 定义与范围界定”
• “功能优先级排序方法”
• “设计冲刺引导”
• “用于产品路线图的影响地图”