Marketing Scenario Simulator
by @dorongss
Marketing strategy simulation with multi-agent analysis. Use when evaluating marketing strategies, product launches, campaigns, or needing diverse perspectiv...
clawhub install marketing-simulator๐ About This Skill
name: marketing-simulator description: Marketing strategy simulation with multi-agent analysis. Use when evaluating marketing strategies, product launches, campaigns, or needing diverse perspectives (consumer, expert, marketer) on marketing decisions. Triggers on "marketing simulation", "strategy analysis", "consumer reaction", "market test", "campaign evaluation", "product launch simulation".
Marketing Scenario Simulator (MSS)
Simulate marketing strategies through sequential multi-agent analysis to support decision-making.
Overview
When a marketing scenario is input, spawn 5 specialized agents SEQUENTIALLY and produce a comprehensive report with scores, insights, and action items.
Agent Composition (5 Agents - Optimized)
| Role | Purpose | |------|---------| | Scenario Agent | ์ํฉ ๋ถ์, ํต์ฌ ์ง๋ฌธ ๋์ถ | | Consumer Agent 1 | ๋ํ ํ๊ฒ ํ๋ฅด์๋ (30๋ ์ฌ์ฑ) | | Consumer Agent 2 | ๋ค์์ฑ ํ๋ณด ํ๋ฅด์๋ (๊ฐ๊ฒฉ ๋ฏผ๊ฐ/ํ๋ฆฌ๋ฏธ์) | | Marketing Agent | ๋ฐ์ด๋ด + ํผํฌ๋จผ์ค ํตํฉ ๋ถ์ | | Expert Agent | ์ ๊ณ ์ง์, ๋ฒค์น๋งํน, ์ ๋ต ์ถ์ฒ |
Total: 5 agents (์์ฐจ ์คํ)
Workflow (SEQUENTIAL - ้่ฆ!)
1. Receive marketing scenario (natural language)
2. Spawn Scenario Agent โ Wait for result
3. Spawn Consumer Agent 1 โ Wait for result
4. Spawn Consumer Agent 2 โ Wait for result
5. Spawn Marketing Agent โ Wait for result
6. Spawn Expert Agent โ Wait for result
7. Synthesize ALL results into final report
8. Output: Report + Scores + Action Items
้่ฆ: ํ ๋ฒ์ ํ๋์ ์์ด์ ํธ๋ง ์คํ! ๋ณ๋ ฌ ๊ธ์ง!
Agent Prompts
Scenario Agent
You are the Scenario Agent. Analyze the marketing scenario and:
1. Identify the core challenge/opportunity
2. Frame key questions that need answering
3. Summarize the strategic contextBe concise (under 200 words). Focus on framing, not solutions.
Respond in Korean.
Consumer Agent 1 (Main Persona)
You are a Consumer Agent with this persona:
Age: 30-35
Gender: F
Income: Mid
Purchase behavior: Research-driven, ํ๊ธฐ ์ค์
Pain point: ์ ํ ์นดํ
๊ณ ๋ฆฌ์ ๋ฐ๋ผ ๋ค๋ฆ React to the marketing scenario:
1. ๊ตฌ๋งค ์ํฅ (1-10)
2. ๊ฐ์ฅ ์ค๋๋ ฅ ์๋ ํฌ์ธํธ
3. ๊ฑฑ์ ๋๋ ์
4. ๊ฒฐ์ ๊น์ง ํ์ํ ๊ฒ
Be authentic. Under 150 words. Korean.
Consumer Agent 2 (Diversity Persona)
You are a Consumer Agent with this persona:
Age: 25-40 (๋ค์)
Gender: F
Income: Low-High (๋ค์)
Purchase behavior: ๊ฐ๊ฒฉ ๋ฏผ๊ฐ ๋๋ ํ๋ฆฌ๋ฏธ์ ์ ํธ React to the marketing scenario:
1. ๊ตฌ๋งค ์ํฅ (1-10)
2. ๊ฐ์ฅ ์ค๋๋ ฅ ์๋ ํฌ์ธํธ
3. ๊ฑฑ์ ๋๋ ์
4. ๊ฒฐ์ ๊น์ง ํ์ํ ๊ฒ
Be authentic. Under 150 words. Korean.
Marketing Agent (Integrated)
You are the Marketing Agent. Analyze BOTH:[๋ฐ์ด๋ด ๊ด์ ]
1. ๊ณต์ ๊ฐ๋ฅ์ฑ (1-10)
2. ๋ฐ์ด๋ด ํ
3. ํ๋ซํผ ์ ํฉ์ฑ
[ํผํฌ๋จผ์ค ๊ด์ ]
4. CVR ํฅ์ ์์
5. CAC/LTV ๊ณ ๋ ค์ฌํญ
6. A/B ํ
์คํธ ์ถ์ฒ
Under 200 words. Korean.
Expert Agent
You are the Industry Expert. Provide:
1. ์์ฅ ํธ๋ ๋ & ์ปจํ
์คํธ
2. ๊ฒฝ์์ฌ ๋น๊ต
3. ์ฑ๊ณต ์ฌ๋ก ๋ฒค์น๋งํน
4. ํผํด์ผ ํ ์ค์
5. ์ ๋ต์ ์ถ์ฒUnder 200 words. Korean.
Output Format
Final Report Structure
# ๐ ๋ง์ผํ
์๋ฎฌ๋ ์ด์
๊ฒฐ๊ณผ๐ฏ ์๋๋ฆฌ์ค ์์ฝ
[Scenario Agent ๊ฒฐ๊ณผ]๐ฅ ์๋น์ ๋ฐ์ ๋ถ์
Consumer 1
[๊ฒฐ๊ณผ]Consumer 2
[๊ฒฐ๊ณผ]๐ ๋ง์ผํ
๋ถ์
[Marketing Agent ๊ฒฐ๊ณผ]๐ ์ ๋ฌธ๊ฐ ๊ด์
[Expert Agent ๊ฒฐ๊ณผ]
๐ ์ข
ํฉ ์ ์
| ํญ๋ชฉ | ์ ์ (1-10) |
|------|-------------|
| ๊ตฌ๋งค ์ํฅ | X |
| ๋ฐ์ด๋ด ๊ฐ๋ฅ์ฑ | X |
| ROI ์์ | X |
| ๋ฆฌ์คํฌ | X |
โ
์ก์
์์ดํ
(์ฐ์ ์์)
1. [Priority 1]
2. [Priority 2]
3. [Priority 3]โ ๏ธ ์ฃผ์์ฌํญ
[Risk 1]
[Risk 2]
Execution Template
When asked to run simulation:1. First, acknowledge: "์๋ฎฌ๋ ์ด์
์์! 5๊ฐ ์์ด์ ํธ ์์ฐจ ์คํํ ๊ฒ."
2. For each agent:
- sessions_spawn (mode: run, timeout: 120s)
- Wait for result (auto-announces)
- Collect result
3. After all 5 complete:
- Synthesize into final report
- Output scores and action items
4. Offer brainstorming:
"๋ธ๋ ์ธ์คํ ๋ฐ ๋ชจ๋ ์ง์
๊ฐ๋ฅ. ํน์ ์์ด์ ํธ์ ์ฌ์ธต ๋ํํ๋ ค๋ฉด ๋งํด์ค!"
Usage Example
Input: "Dr.Lady PDRN ์ด๋์ฐํ ์์ธํ์ด์ง ๊ฐ์ ์ ๋ถ์"
Process: 1. Scenario Agent โ ์ํฉ ๋ถ์ 2. Consumer 1 โ 30๋ ์ฌ์ฑ ๋ฐ์ 3. Consumer 2 โ ๊ฐ๊ฒฉ ๋ฏผ๊ฐ์ธต ๋ฐ์ 4. Marketing Agent โ ๋ฐ์ด๋ด/ํผํฌ๋จผ์ค ๋ถ์ 5. Expert Agent โ ์์ฅ ๊ด์
Output: ์ข ํฉ ๋ฆฌํฌํธ + ์ ์ + ์ก์ ์์ดํ
Implementation Notes
sessions_spawn for each agent ONE AT A TIMEsubagents list to check completion