用 n8n + GPT-4 Mini 清洗 Google Sheet 中的公司名，提升冷邮件个性化效果

输入：含 Company 列的 Google Sheet URL → 处理：LLM 自动去除后缀、标语与格式问题 → 输出：新增 Cleaned Companies 工作表，含清洗后的公司名

给谁

负责冷邮件外联的销售或市场运营人员

你会得到

将 messy 公司名（如 "Techwave Solutions LLC"）清洗为自然、专业的简称（如 "Techwave"），降低模板感，提升回复率

你需要

n8n · google-sheets

步骤

1. 配置凭证并准备 Google Sheet

   在 n8n 中添加 Google Sheets OAuth2 和 OpenAI API 凭证。创建一个新的 Google Sheet，确保第一行表头包含 "Company" 列，下方填入待清洗的 leads 数据。

2. 触发工作流并创建目标工作表

   运行工作流后，通过 n8n Form 提交 Sheet URL。工作流会自动在原 Spreadsheet 中创建一个名为 "Cleaned Companies" 的新工作表（绿色标签），用于存放结果。

3. 读取线索数据

   工作流从提交 URL 对应的首个工作表（sheet ID "0"）中读取全部 leads 数据。

4. 用 LLM 清洗公司名

   将 Company 列逐条传入 LLM Chain。模型使用 GPT-4.1-mini（可在 n8n 中切换为 gpt-4.1 或 gpt-4），配合结构化输出解析器，按 system prompt 规则清洗：去除 LLC/Inc. 等后缀、Systems/Solutions 等冗余词、标语与描述短语，应用 title case，保留可识别性。批次大小为 5。

   运行提示词 · 来源原文。Clean Company Name 节点的 messages.messageValues[0].message（system prompt）

   复制提示词

   **Situation** 
   You are processing a lead list for a cold email campaign where company names  will be used as personalization variables. The raw company names in the dataset contain various formatting inconsistencies, unnecessary suffixes, taglines, and casing issues that make them unsuitable for professional, natural, personalized outreach.  
   
   **Task** 
   Clean and standardize company names from a lead list to create professional, polished variables suitable for cold email personalization. For each company name provided, you must: 
   1. Remove common business suffixes (LLC, Inc., Corp., Corporation, Ltd., Limited, Co., Company, etc.) 
   2. Remove common words that are in the formal company name but that a human likely wouldn't include when identifying the company in an email (Systems, Solutions, Communications, Agency, Group, Partners, etc.)
   3. Remove taglines, slogans, or descriptive phrases (typically found after dashes, pipes, or "the" statements) 
   4. Apply proper title case formatting 
   5. Shorten names to their most recognizable form when possible while maintaining professionalism 
   6. Ensure the result maintains the company's professional identity and recognizability  
   
   **Objective** 
   Transform formal, cluttered company names into clean, professional versions that will make cold emails feel more natural, personal and polished, ultimately improving engagement rates and response likelihood in the outreach campaign.
   
   **Knowledge** 
   - Business suffixes to remove include: LLC, Inc., Corp., Corporation, Ltd., Limited, Co., Company, LP, LLP, PC, PLLC, and their variations with different punctuation 
   - Taglines often appear after symbols like: dash (-), pipe (|), colon (:), or phrases starting with "the," "your," "we," etc. 
   - Proper title case means capitalizing the first letter of each major word, but keeping articles, prepositions, and conjunctions lowercase unless they start the name. Acronyms should remain in all caps.
   - Maintain professional tone 
   - Keep established business names recognizable (e.g., "Microsoft Corporation" becomes "Microsoft") 
   - Avoid over-shortening if it makes the company unrecognizable or unprofessional 
   - Service descriptions and taglines should be completely removed (e.g., "ExtraSlice, LLC - Workspace as a Service" becomes "ExtraSlice")  
   
   <examples> 
   Input: "ExtraSlice, LLC - Workspace as a Service" 
   Output: "ExtraSlice"  
   
   Input: "ACRE Solutions" 
   Output: "ACRE"  
   
   Input: "MRP Capital Group" 
   Output: "MRP"  
   
   Input: "Paul & Associates Real Estate Services" 
   Output: "Paul & Associates"  
   
   Input: "Global Valuation & Consulting, LLC." 
   Output: "Global Valuation & Consulting"  
   
   Input: "Agent Career Education | Ace" 
   Output: "Agent Career Education"  
   
   Input: "PROPERTY AGENTS" 
   Output: "Property Agents"  
   
   Input: "Agently - Agent Success Platform for Brokerages" 
   Output: "Agently"  
   
   Input: "Monserrate Premium Home Agents" 
   Output: "Monserrate"  
   
   Input: "Belmont Howard Group, LLC."
   Output: "Belmont Howard"  
   
   Input: "BruZiv Partners, LLC" 
   Output: "BruZiv"  
   
   Input: "Van Vlissingen and Co." 
   Output: "Van Vlissingen" 
   </examples>
   
   It's critical that the company's professional recognizability while removing all unnecessary suffixes and taglines - the cleaned name must sound natural and professional when used in a business email context.

5. 合并原始数据并保存结果

   将 LLM 输出的 cleanCompanyName 与原始 lead 数据按位置合并，添加 "Clean Company Name" 列，排除中间字段后，追加写入 "Cleaned Companies" 工作表。

你正在复现 n8n 模板 #5647：用 GPT-4 Mini 清洗 Google Sheet 中的公司名。请先确认以下前置条件：1) n8n 实例已就绪；2) Google Sheets OAuth2 和 OpenAI API 凭证已配置；3) 目标 Sheet 包含 "Company" 列。然后导入 workflow.json 或按步骤手动搭建：表单触发 → 创建目标 sheet → 读取 leads → LLM Chain 清洗（含上述 system prompt）→ 合并 → 保存。建议先用小批量测试，验证输出格式后再全量运行。