🚨 你唔係效率低，你係冇系統
你有冇試過：
- 每次接新 client，都要重新 research
- 開十幾個分頁搵資料
- 翻舊 proposal copy + paste
- 改名、改日期、改內容
- 最後用咗 3–5 個鐘先出到一份 proposal
問題唔係你慢，而係你每次都由零開始。
真正高效唔係做得快，而係：
「你有冇一條可以重複用嘅 workflow。」
🎯 呢篇文章會俾你咩？
呢篇係一個完整資源包，你可以直接拎去用：
- NotebookLM Client Research Prompt
- Insight Database（Google Sheets）結構
- Proposal 自動生成 Prompt
- 完整 AI Workflow Pipeline
（資源整理來源：）
🧠 Part 1：Client Research Prompt（NotebookLM）
🧩 用途
將 client 資料（PDF / 網站 / notes）轉成：
- 結構化 research report
- 有來源支持（避免 AI 幻覺）
- 可直接用於 proposal
🧾 Prompt（直接複製用）
Based on the sources in this NotebookLM notebook, generate a structured client research report.
The report should include the following sections:
1. Client Summary
- Company overview
- Current hiring context
- Employer branding positioning
2. Competitor Landscape
- Identify 2–3 main competitors mentioned in the sources
- Summarize their employer branding positioning
- Highlight key differences compared with the client
3. Audience Pain Points
- Extract the main concerns of mid-career professionals in Hong Kong
- Identify common frustrations candidates have during hiring
- Summarize what candidates value most when evaluating employers
4. Recommended Workshop Angles
Based on the analysis above, suggest 3 strategic workshop angles that could help the client improve their employer branding and hiring communication.
Format the report clearly with section headings and bullet points.
Only use information grounded in the provided sources. Reference the sources where relevant.
💡 用法
- 將 client 資料放入 Google Drive
- 匯入 NotebookLM
- 使用以上 prompt
- 輸出 standardized research report
「AI 最有價值唔係生成內容，而係整理你已有嘅資料。」
📊 Part 2：Insight Database（Google Sheets）
❌ 常見問題
- 每個 client 一份 Doc
- 用完就算
- Insight 無法累積
✅ 解法：建立 Insight Database
每個 client → 一行資料
📋 Sheet 結構（直接用）
Client Name
Industry
Team Size
Core Hiring Challenge
Employer Branding Maturity
Key Audience
Competitor Notes
Recommended Workshop Angle
Audience Pain Points
Common Sales Objections
Tags
（結構來源：）
🤖 Research → Data Prompt
Based on the research report from the previous step, extract structured fields for an insight database.
Return the following fields:
Client name
Industry
Team size
Core hiring challenge
Employer branding maturity
Competitor notes
Recommended workshop angle
Audience pain points
Common sales objections
Return the result in structured format suitable for inserting into a Google Sheets row.
💡 核心價值
當你開始累積 database，你可以：
- 比較唔同行業
- 發現重複 pain points
- 優化 proposal messaging
- 建立自己嘅 consulting framework
「文件會消失，但資料會複利。」
🧾 Part 3：Proposal 自動生成 Prompt
🎯 用途
將以下資料：
- Client brief
- Research report
- Insight database
轉成：
👉 一份完整 proposal（webpage / doc）
🧠 Prompt（直接用）
Using the client brief, and research summary, generate a structured proposal webpage.
The webpage should be written in a clear consulting style and organized into sections.
Structure the webpage with the following sections.
Section 1 — Proposal Header
Include:
• Proposal title
• Client name
• Engagement topic
• Short introductory sentence
Section 2 — Executive Summary
Summarize the situation and opportunity.
Include:
• Current hiring situation
• Key communication challenges
• Why this workshop is recommended
Section 3 — Client Hiring Challenges
Include:
• Hiring communication inconsistencies
• Employer branding clarity
• Candidate experience issues
Section 4 — Market & Talent Landscape
Include:
• What mid-career professionals care about
• Hiring expectations
• Employer branding trends
Section 5 — Competitor Landscape
Include:
• Competitor positioning
• Messaging differences
• Opportunities
Section 6 — Recommended Workshop Design
Break into sessions and objectives
Section 7 — Expected Outcomes
Section 8 — Implementation Roadmap
Section 9 — Next Steps
Write as a clean webpage layout using headings and concise paragraphs.
Ensure all insights are grounded in the provided research sources and insight database.
（來源：）
💡 使用重點
- 用固定 structure（保持一致）
- 加入 brand tone（提升質感）
- 用 database insight（提升命中率）
⚙️ 完整 Workflow Pipeline
🔁 架構
1. Google Drive（放資料）
↓
2. NotebookLM（做 research）
↓
3. AI（生成 report）
↓
4. Human Review（檢查）
↓
5. AI（extract data）
↓
6. Google Sheets（累積 insight）
↓
7. AI（生成 proposal）
🧠 核心邏輯
大部分人用 AI：
👉 一問一答
真正有效嘅做法：
👉 多 source → 多步驟 → 一個高質 output
🛠️ 實戰流程（照做版）
Step 1：建立 Client Folder
放：
- 公司資料
- 官網內容
- competitor
Step 2：NotebookLM 做 Research
用固定 prompt → 輸出 report
Step 3：人工 Review（必做）
確保：
- 無 hallucination
- 無錯誤資料
Step 4：轉成 Database
用 extract prompt → 寫入 Google Sheets
Step 5：生成 Proposal
用 final prompt → 自動輸出
⚠️ 常見錯誤
- 跳過 human review
- 無建立 database
- 每次改 prompt（無標準化）
- 無使用過往 insight
🧠 最重要一句
當你有呢條 pipeline：
- 每個 client 都會變 asset
- 每份 proposal 都會更準
- 每次 workflow 都會更快
你唔再係：
👉 用時間換收入
而係：
👉 用系統放大產出
.jpg)
