用 n8n 表单访谈生成商业模式画布信息图

用户填写四阶段网页表单 → n8n 聚合格式化 → AI Agent 生成 9 要素 BMC → Gemini 输出专业信息图

给谁

需要在工作坊或教学中快速生成商业模式画布（BMC）的顾问、培训师、加速器运营者，或任何运行 n8n 的人

你会得到

受访者通过网页表单回答商业构想问题，系统自动生成包含 9 要素的商业模式画布文本及专业信息图

你需要

n8n

步骤

1. 配置并发布四阶段访谈表单

   在 n8n 中导入模板后，检查并调整表单节点。表单分为 4 个阶段： - Phase 1（核心商业概念）：目标客户、痛点、现有应对方式、价值主张、自身优势与动机 - Phase 2（客户触点）：信息获取渠道、首次接触方式、购买/注册渠道、客户关系类型、复购与推荐机制 - Phase 3（交付手段）：关键活动、可 leveraged 资源、缺失资源/合作方 - Phase 4（资金流动）：收入模式、增长策略、价格区间、成本结构 每个阶段对应一个 n8n Form 节点（On form submission → Form2 → Form3 → Form4），按顺序连接。可根据需要修改字段标签与占位文本。激活工作流后，分享 Production URL 给最终用户。

2. 聚合并格式化访谈数据

   添加或确认 'Format Interview Responses' Code 节点。该节点将前四个表单节点的输出合并为统一数据结构，按阶段分组，过滤掉 submittedAt、formMode 等系统字段，生成供 AI Agent 读取的格式化问答列表。

   运行提示词 · 来源原文。'Format Interview Responses' Code 节点的 JS Code 字段

   复制提示词

   const phase1 = $('On form submission').item.json;
   const phase2 = $('Form2').item.json;
   const phase3 = $('Form3').item.json;
   const phase4 = $('Form4').item.json;
   
   const allData = { ...phase1, ...phase2, ...phase3, ...phase4 };
   
   const formatted = [
     { phase: "Grasping the Core (Most Important)", data: Object.entries(phase1) },
     { phase: "Customer Touchpoints (External Environment)", data: Object.entries(phase2) },
     { phase: "Means of Delivery (Internal Environment)", data: Object.entries(phase3) },
     { phase: "Money Flow", data: Object.entries(phase4) },
   ];
   
   return [{ json: { allData, formatted } }];

3. 配置 AI Agent 生成 9 要素 BMC

   配置 'AI Canvas Generator' AI Agent 节点： 1. 将 'AWS Bedrock Chat Model'（或你偏好的 LLM）连接到该 Agent 的 Language Model 端口。 2. 在 Agent 的 Text 字段填入用户问答模板（见下方提示词），用于将格式化后的表单数据注入对话。 3. 在 Options → System Message 中填入系统提示词（见下方提示词），定义验证规则与 9 要素输出格式。 AI 会先验证输入是否相关、完整、连贯；若无效则返回 ERROR JSON；若有效则生成 9 要素文本。

   运行提示词 · 来源原文。'AI Canvas Generator' 节点的 Text 字段（用户输入模板）

   复制提示词

   =user's answers:
   
   {{ 
   $json.formatted
     .map(phase => 
       `## ${phase.phase}\n\n` + 
       phase.data
         .filter(([q]) => !['submittedAt', 'formMode'].includes(q))
         .map(([q, a]) => `Q: ${q}\nA: ${a}`)
         .join('\n\n')
     )
     .join('\n\n---\n\n') 
   }}

   运行提示词 · 来源原文。'AI Canvas Generator' 节点的 Options → System Message 字段

   复制提示词

   =You are an expert business consultant. Based on the user's answers regarding their business idea, your task is to organize and create the 9 elements of the Business Model Canvas.
   
   ## Step 1: Validation (REQUIRED)
   
   Before generating the Business Model Canvas, you MUST validate the user's answers:
   
   ### Validation Criteria
   1. **Relevance Check**: Answers must be related to business/service concepts
   2. **Completeness Check**: Key questions (Customer, Value Proposition, Revenue) must have meaningful answers
   3. **Coherence Check**: Answers should be logically consistent with each other
   
   ### Invalid Answer Examples
   - Random text, gibberish, or test data (e.g., "asdf", "test", "xxx")
   - Completely unrelated content (e.g., recipes, song lyrics, jokes)
   - Contradictory information that cannot form a coherent business model
   - Answers that are too vague to extract any business insight (e.g., "something", "anything")
   
   ### Validation Result
   
   If answers are INVALID, respond ONLY with this JSON format and nothing else:
   ```json
   {
     "status": "ERROR",
     "reason": "Brief explanation of what was wrong",
     "message": "We were unable to create a Business Model Canvas from your responses. Please provide specific information about your business idea and try again."
   }
   ```
   
   If answers are VALID, proceed to Step 2.
   
   ---
   
   ## Step 2: Generate Business Model Canvas
   
   Only proceed here if validation passed.
   
   ### Rules for Elaboration
   Do not simply transcribe the user's answers. Instead, slightly flesh out the content based on the following perspectives:
   
   1. **Concretize**:
      - If there are vague expressions, infer from the context and add specific details.
      - *Example:* Change "SNS" to specific platforms like "Instagram" or "X" based on the target audience.
   
   2. **Complement**:
      - Add elements that are clearly necessary from the context, even if not explicitly stated in the answers.
      - *Example:* Organize Channels into the flow of "Awareness -> Purchase -> Delivery".
   
   3. **Constraints**:
      - Do not interpret in a way that deviates from the user's intent.
      - Do not over-expand the content (keep additions light and relevant).
      - If you add parts based on speculation, mark them as "(Assumed)" if necessary.
   
   ### Output Format
   
   Extract and organize the information into the following 9 elements:
   
   **【Customer Segments (CS)】**
   * Who are we creating value for?
   
   **【Value Propositions (VP)】**
   * What problems are we solving? What value do we deliver to the customer?
   
   **【Channels (CH)】**
   * How do we reach the customer? (Organize by Awareness -> Purchase -> Delivery)
   
   **【Customer Relationships (CR)】**
   * What type of relationship do we establish with the customer?
   
   **【Revenue Streams (RS)】**
   * How does the business generate income?
   
   **【Key Resources (KR)】**
   * What key assets are required? (Skills, qualifications, equipment, etc.)
   
   **【Key Activities (KA)】**
   * What key activities are necessary to deliver the Value Propositions?
   
   **【Key Partners (KP)】**
   * Who are the external partners or suppliers?
   
   **【Cost Structure (CS)】**
   * What are the major costs involved? (Fixed costs, variable costs, initial investment)

4. 验证输出并路由错误

   添加或确认 'Validate Canvas Output' Code 节点与 'If_is_error' 分支节点： - Code 节点检查 AI Agent 输出是否包含 ERROR 状态；若含错误，提取 message 字段并标记 isError=true。 - If 节点判断 isError：为 true 时导向错误结束页，为 false 时将 canvas 文本传递给图像生成节点。 错误页向用户显示友好提示，要求重新填写；成功分支继续执行。

   运行提示词 · 来源原文。'Validate Canvas Output' Code 节点的 JS Code 字段

   复制提示词

   const output = $('AI Canvas Generator').item.json.output;
   
   if (output.includes('"status": "ERROR"') || output.includes('"status":"ERROR"')) {
     const errorMatch = output.match(/"message":\s*"([^"]+)"/);
     const errorMessage = errorMatch ? errorMatch[1] : 'There was an issue with your input. Please start again from the beginning.';
     
     return [{
       json: {
         isError: true,
         errorMessage
       }
     }];
   }
   
   return [{
     json: {
       isError: false,
       canvas: output
     }
   }];

5. 生成信息图并交付结果

   配置 'Generate an image' Google Gemini 节点： 1. 绑定 Google Gemini API 凭证（googlePalmApi 类型）。 2. 选择支持图像生成的模型，如 models/gemini-3-pro-image-preview。 3. 在 Prompt 字段填入下方提示词，该提示词将前序节点的 canvas 文本注入，要求 Gemini 按标准 9 宫格布局生成专业信息图。 成功时，'Completed' 表单结束页向用户展示生成的图像链接；失败时已在第 4 步被拦截。

   运行提示词 · 来源原文。'Generate an image' 节点的 Prompt 字段

   复制提示词

   =## Business Model Canvas text
   {{ $json.canvas }}
   
   ## Visual Output Request
   
   Based on the organized Business Model Canvas text above, generate a professional **infographic image**.
   
   **【Required 9 Elements】**
   The image MUST include all 9 Business Model Canvas elements:
   1. Key Partners (KP)
   2. Key Activities (KA)
   3. Key Resources (KR)
   4. Value Propositions (VP)
   5. Customer Relationships (CR)
   6. Channels (CH)
   7. Customer Segments (CS)
   8. Cost Structure (CS)
   9. Revenue Streams (RS)
   
   **【Visual Style & Format Specifications】**
   1. **Layout Rigidity**: Strictly adhere to the standard Business Model Canvas 9-block grid layout:
      - Left column: Key Partners
      - Center-left column (split): Key Activities (top), Key Resources (bottom)
      - Center column: Value Propositions
      - Center-right column (split): Customer Relationships (top), Channels (bottom)
      - Right column: Customer Segments
      - Bottom-left: Cost Structure
      - Bottom-right: Revenue Streams
   
   2. **Aesthetic Style**: Emulate a clean, professional, digital infographic design, similar to high-quality BMC templates used in corporate presentations.
   
   3. **Visual Elements**:
      * Include minimalist, relevant **icons** next to each block title to enhance visual clarity
      * Use professional header colors for each block's title area (e.g., muted blues or greys as accents against a white background)
      * Ensure the dividing lines between blocks are sharp and clear
   
   4. **Text Rendering & Placement**:
      * Render the summarized key points legibly *within* their corresponding block boundaries
      * **Crucial**: Text must be high-contrast and must not overflow or overlap with borders
      * Do not overcrowd the blocks; use bullet points for readability
   
   **【Output Goal】**
   A single, high-resolution image that visualizes the entire business model at a glance, with all 9 elements clearly visible and properly positioned.

你需要一个 n8n 实例、AWS Bedrock 凭证（默认 Claude，可替换）以及 Google Gemini API 密钥。导入模板后，先配置两个凭证节点，再按需调整表单问题，最后激活工作流并分享 Production URL。最终用户无需注册，直接在浏览器中完成 4 阶段表单即可获得 BMC 信息图。