AgentSkillsCN

variation-matrix

基于模块化组件（钩子、正文、CTA），按照优先级排序进行测试，生成所有可能的广告变体。当您需要规划创意测试、最大化现有素材价值，或构建系统化的变体测试日程时，可选用此方法。

SKILL.md
--- frontmatter
name: variation-matrix
description: Generate all possible ad variations from modular components (hooks, bodies, CTAs) with prioritized testing order. Use when planning creative tests, maximizing existing assets, or building a systematic variation testing schedule.

Variation Matrix

Create all possible ad combinations from modular components.

Process

Step 1: Input Available Components

Hooks (H) List all available hook clips/copy:

  • •H1: [Description]
  • •H2: [Description]
  • •H3: [Description] ...

Bodies (B) List all available body sections:

  • •B1: Testimonial body
  • •B2: Explainer body
  • •B3: Social proof compilation ...

CTAs (C) List all available call-to-action clips/copy:

  • •C1: Direct CTA
  • •C2: Scarcity CTA
  • •C3: Story-based CTA ...

Step 2: Create Combination Matrix

Total Variations = H x B x C

Example: 5 hooks x 3 bodies x 2 CTAs = 30 variations

code
| Variation | Hook | Body | CTA |
|-----------|------|------|-----|
| V1        | H1   | B1   | C1  |
| V2        | H1   | B1   | C2  |
| V3        | H1   | B2   | C1  |
| V4        | H1   | B2   | C2  |
| V5        | H1   | B3   | C1  |
...

Step 3: Prioritize by Predicted Performance

Scoring Factors:

  • •Hook strength (past performance or score)
  • •Body relevance to hook
  • •CTA match to body
  • •Overall flow/coherence

Priority Tiers:

Tier 1 - Test First (High confidence)

  • •Best hooks + proven bodies
  • •Logical flow combinations
  • •Based on winning patterns

Tier 2 - Test Second (Medium confidence)

  • •Good hooks + new bodies
  • •Testing new combinations
  • •Iterating on Tier 1 learnings

Tier 3 - Test Later (Experimental)

  • •Untested combinations
  • •Unusual pairings
  • •"What if" variations

Step 4: Suggest Testing Order

Phase 1: Prove the hooks Test all hooks with best body + best CTA

  • •V1, V5, V9, V13, V17 (one per hook)

Phase 2: Find best body for winning hooks Take top 2-3 hooks, test all bodies

  • •Winning hooks x all bodies

Phase 3: Optimize CTA Take top combinations, test all CTAs

Phase 4: Full matrix test Roll out remaining combinations based on learnings

Step 5: Output Numbered Variation List

code
## VARIATION MATRIX: [Offer/Campaign]

### COMPONENT INVENTORY

**HOOKS ([#] available)**
| ID | Description | Score | Source |
|----|-------------|-------|--------|
| H1 | [Description] | 9/10 | [Clip/Copy] |
| H2 | [Description] | 8/10 | [Clip/Copy] |
...

**BODIES ([#] available)**
| ID | Type | Description | Duration |
|----|------|-------------|----------|
| B1 | Testimonial | [Description] | 20s |
| B2 | Explainer | [Description] | 25s |
...

**CTAs ([#] available)**
| ID | Type | Description |
|----|------|-------------|
| C1 | Direct | "Click the link below" |
| C2 | Scarcity | "Sale ends Friday" |
...

---

### FULL MATRIX

Total Possible Variations: [#]

| # | Hook | Body | CTA | Priority | Est. Length |
|---|------|------|-----|----------|-------------|
| 1 | H1 | B1 | C1 | Tier 1 | 30s |
| 2 | H1 | B1 | C2 | Tier 2 | 30s |
| 3 | H1 | B2 | C1 | Tier 2 | 35s |
...

---

### TESTING PHASES

**PHASE 1: Hook Test (Week 1)**
Variations: V1, V6, V11, V16, V21
Budget: $[X]
Goal: Identify top 2-3 hooks

| Var | Components | Priority |
|-----|------------|----------|
| V1 | H1+B1+C1 | Test |
| V6 | H2+B1+C1 | Test |
...

**PHASE 2: Body Test (Week 2)**
Take winning hooks, test all bodies
Budget: $[X]
Goal: Find best body for each hook

**PHASE 3: CTA Optimization (Week 3)**
Take winning combinations, test CTAs
Budget: $[X]
Goal: Maximize conversion

**PHASE 4: Scale Winners (Week 4+)**
Full variations of proven combinations

---

### QUICK REFERENCE

**Top 10 Priority Variations:**
1. V[X]: H[X] + B[X] + C[X] - [Rationale]
2. V[X]: H[X] + B[X] + C[X] - [Rationale]
...

**Experimental Combinations:**
- V[X]: [Why worth testing despite lower confidence]

---

### PRODUCTION CHECKLIST

**To Create [#] Variations:**
- [ ] Hook clips ready: [#]
- [ ] Body sections ready: [#]
- [ ] CTA clips ready: [#]
- [ ] Editing timeline: [Date]
- [ ] Launch date: [Date]

**Naming Convention:**
[Offer]_[Hook#]_[Body#]_[CTA#]_v[version]
Example: Rise_H1_B2_C1_v1

Variation Strategy (LeadsIcon)

Modular Thinking:

  • •Each component has a job
  • •Hook: Grab attention
  • •Body: Build belief
  • •CTA: Remove friction

Testing Efficiency:

  • •Don't test everything at once
  • •Isolate variables
  • •Build on winners
  • •Kill losers fast

From Winners:

  • •Create 10+ variations of each winner
  • •Test new hooks on winning bodies
  • •Test new bodies with winning hooks
  • •Compound improvements

Source: LeadsIcon, VISCAP