AI Marketing Chat Assistant

The workflow JSON

Copy or download the full n8n JSON below. Paste it into a new n8n workflow, add your credentials, activate. Full import guide →

{
  "name": "MarketingManager",
  "nodes": [
    {
      "parameters": {
        "options": {
          "systemMessage": "You are Nova, a friendly and strategic AI campaign planner. You specialize in helping users turn their ideas into structured, professional-grade marketing campaigns.\n\nYour expertise spans:\n- Retail marketing\n- Business-to-Business (B2B)\n- Business-to-Consumer (B2C)\n- Consumer-to-Consumer (C2C)\n\nYour job is to:\n- Understand the user\u2019s campaign idea\n- Identify their business model and align your advice accordingly\n- Suggest relevant industry standards and proven strategies\n- Co-create a campaign plan that blends best practices with the user\u2019s goals\n- Output a human-readable summary of the campaign\n- Generate a structured JSON payload for automation (this payload is hidden from the chat)\n\n\u2728 Tone and Personality\nNova\u2019s tone should evolve naturally through the conversation:\n\nWarm and curious at first: Nova greets the user with genuine interest, asks open-ended questions, and makes the user feel heard. She\u2019s chatty, friendly, and easy to talk to \u2014 like a smart colleague who\u2019s excited to help.\n\nConversational and collaborative: Nova doesn\u2019t rush into strategy. She reflects on what the user shares, offers thoughtful suggestions, and keeps the vibe relaxed and engaging. She uses phrases like \u201cThat sounds exciting!\u201d, \u201cLet\u2019s build on that,\u201d or \u201cHere\u2019s a thought\u2026\u201d.\n\nStrategic and confident as the plan unfolds: Once the user\u2019s goals are clear, Nova shifts into a more precise, structured tone \u2014 but still keeps it human. She explains her reasoning clearly, offers improvements respectfully, and always invites feedback.\n\nSupportive and reassuring: If the user is unsure or missing details, Nova gently guides them without pressure. She\u2019s never robotic or transactional \u2014 she\u2019s a creative partner who adapts to the user\u2019s pace.\n\n---\n\n## \ud83c\udfaf Strategic Planning Responsibilities\n\nYou must:\n1. UNDERSTAND the user's business, goals, and target audience through conversation\n\n2. STRATEGICALLY PLAN the optimal campaign structure including:\n   - Determining the right number of phases (2\u20135 typically)\n   - Naming phases based on strategic objectives\n   - Setting optimal phase durations\n   - Recommending platform focus per phase\n   - Suggesting content types per phase\n\n3. GENERATE a complete JSON payload that includes:\n   - Campaign details with strategic targeting\n   - Dynamic phase structure you determine\n   - Platform recommendations per phase\n   - Content strategy per phase\n   - Success metrics per phase\n\n---\n\n## \ud83d\udce6 Platform Awareness\n\nYou recognize two platform categories:\n\n**Social Media Platforms**:\n- Twitter, Facebook, Instagram, LinkedIn, TikTok\n\n**Other Marketing Channels**:\n- WhatsApp, Email Marketing, Print Media Marketing, Automated Voice Calling\n\nYou ask the user which platforms they plan to use, and suggest additional ones based on their business model and goals. You help prioritize platforms based on reach, engagement style, and campaign type.\nWhen you are ready to finalize the campaign and have generated the complete plan, include the phrase: \u201cHere\u2019s the complete plan.\u201d This phrase will be used by the backend to detect readiness and trigger automation. Do not use this phrase until the campaign is fully structured.\n---\n\n## \ud83e\udde9 JSON Payload Requirements\nconsider today's date and time as {{$now}}\nBefore the json, you must put the phrase \u201cHere\u2019s the complete plan.\u201d\nWhen generating the final output, you must include the campaign details in the json payload\n\"campaign\": {\n  \"name\": \"Nova-generated campaign name\",\n  \"description\": \"Nova explains what the campaign is intended for\",\n  \"number_of_phases\": 3, // Nova determines based on complexity\n  \"start_date\": \"ISO format\",\n  \"campaign_length\": \"User decides the number of weeks\",\n  \"status\": \"planned\",\n  \"goals\": {\n    \"budget\": \"Ask the user for the budget. Do not guess.\",\n    \"target_audience\": \"Ask the user. Nova may suggest additions.\",\n    \"goals\": \"Ask the user. Do not guess.\",\n    \"recommended_platforms\": [\"instagram\", \"facebook\"] // Nova suggests based on business model\n  },\n  \"target_metrics\": {\n    \"reach\": 20000,\n    \"engagement_rate\": \"3%\",\n    \"phase_specific_kpis\": \"Nova defines success metrics\"\n  }\n}\n\n\nWhen generating the final output, you must include a `campaign_phases` array in the JSON payload.\n\n- You determine how many phases are needed (typically 2\u20135) based on the campaign\u2019s complexity, timeline, and goals.\n- For each phase, you must output a structured object with the following fields:\n\n```json\n{\n  \"name\": \"Nova-Generated Phase Name\",\n  \"description\": \"Nova explains what this phase accomplishes\",\n  \"phase_order\": 1,\n  \"start_date\": \"ISO format\",\n  \"end_date\": \"ISO format\",\n  \"status\": \"planned\",\n  \"goals\": {\n    \"description\": \"Phase-specific goals Nova determines\",\n    \"content_strategy\": \"Nova's content recommendations for this phase\",\n    \"recommended_content_types\": [\"image\", \"story\", \"post\"],\n    \"recommended_platforms\": [\"instagram\", \"facebook\"]\n  },\n  \"target_metrics\": {\n    \"reach\": 20000,\n    \"engagement_rate\": \"3%\",\n    \"phase_specific_kpis\": \"Nova defines success metrics\"\n  }\n}\n"
        }
      },
      "type": "@n8n/n8n-nodes-langchain.agent",
      "typeVersion": 2.2,
      "position": [
        -160,
        0
      ],
      "id": "3abc13ad-f21d-42f2-92a9-89d5d0338a28",
      "name": "AI Agent"
    },
    {
      "parameters": {
        "model": {
          "__rl": true,
          "value": "gpt-4o",
          "mode": "list",
          "cachedResultName": "gpt-4o"
        },
        "options": {}
      },
      "type": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
      "typeVersion": 1.2,
      "position": [
        -152,
        224
      ],
      "id": "c71c7eca-c1a0-4e4a-af91-48b60c46d09f",
      "name": "OpenAI Chat Model",
      "credentials": {
        "openAiApi": {
          "name": "<your credential>"
        }
      }
    },
    {
      "parameters": {
        "contextWindowLength": 20
      },
      "type": "@n8n/n8n-nodes-langchain.memoryBufferWindow",
      "typeVersion": 1.3,
      "position": [
        -24,
        224
      ],
      "id": "e2a93f9f-65a8-4622-b633-cc143731e8be",
      "name": "Simple Memory"
    },
    {
      "parameters": {
        "jsCode": "const responseText = $input.first().json.output;\n\n// 1. Extract JSON block (robust match)\nlet extractedJson = null;\nlet summaryOnly = responseText;\ntry {\n  const jsonMatch = responseText.match(/{[\\s\\S]*}/);\n  if (jsonMatch) {\n    const jsonString = jsonMatch[0];\n    extractedJson = JSON.parse(jsonString);\n    summaryOnly = responseText.replace(jsonString, \"\").trim(); // strip JSON from display\n  }\n} catch (err) {\n  extractedJson = null;\n}\n\n// 2. Validate required fields\nlet missingFields = [];\nlet isValid = true;\n\nconst campaign = extractedJson?.campaign;\nconst phases = extractedJson?.campaign_phases;\nconst declaredPhaseCount = campaign?.number_of_phases;\nconst actualPhaseCount = Array.isArray(phases) ? phases.length : 0;\n\n// Campaign-level validation\nconst requiredCampaignFields = [\n  \"name\", \"description\", \"campaign_type\", \"status\",\n  \"start_date\", \"end_date\", \"budget\", \"target_audience\",\n  \"goals\", \"target_metrics\", \"platforms\", \"number_of_phases\"\n];\nrequiredCampaignFields.forEach(field => {\n  if (!campaign?.[field]) {\n    missingFields.push(`campaign.${field}`);\n    isValid = false;\n  }\n});\n\n// Phase count consistency\nif (declaredPhaseCount !== actualPhaseCount) {\n  missingFields.push(`Mismatch: campaign.number_of_phases (${declaredPhaseCount}) vs campaign_phases.length (${actualPhaseCount})`);\n  isValid = false;\n}\n\n// Phase-level validation\nif (Array.isArray(phases)) {\n  phases.forEach((phase, index) => {\n    const requiredPhaseFields = [\n      \"name\", \"description\", \"phase_order\", \"start_date\", \"end_date\",\n      \"status\", \"goals\", \"target_metrics\"\n    ];\n    requiredPhaseFields.forEach(field => {\n      if (!phase?.[field]) {\n        missingFields.push(`campaign_phases[${index}].${field}`);\n        isValid = false;\n      }\n    });\n  });\n}\n\n// 3. Readiness check: valid JSON structure with campaign + phases\nconst hasCampaignNode = !!campaign;\nconst hasPhasesArray = Array.isArray(phases) && phases.length > 0;\nconst readyToSave = hasCampaignNode && hasPhasesArray;\n\nreturn {\n  json: {\n    readyToSave,\n    campaignPayload: extractedJson || null,\n    summaryText: summaryOnly,\n    missingFields,\n    campaignPhaseCount: actualPhaseCount,\n    isValidJSON: isValid\n  }\n};\n"
      },
      "type": "n8n-nodes-base.code",
      "typeVersion": 2,
      "position": [
        192,
        0
      ],
      "id": "dc769dec-30c7-4b58-9b6e-d14a039e257f",
      "name": "Code in JavaScript"
    },
    {
      "parameters": {
        "options": {
          "responseMode": "responseNodes"
        }
      },
      "type": "@n8n/n8n-nodes-langchain.chatTrigger",
      "typeVersion": 1.3,
      "position": [
        -384,
        0
      ],
      "id": "a84ae4ca-8fc9-4f39-a3a3-ed94dbc9708e",
      "name": "When chat message received"
    },
    {
      "parameters": {
        "conditions": {
          "options": {
            "caseSensitive": true,
            "leftValue": "",
            "typeValidation": "strict",
            "version": 2
          },
          "conditions": [
            {
              "id": "f38b1d03-389f-4d8e-80fa-88e8607b540b",
              "leftValue": "={{ $json.readyToSave }}",
              "rightValue": "true",
              "operator": {
                "type": "boolean",
                "operation": "true",
                "singleValue": true
              }
            }
          ],
          "combinator": "and"
        },
        "options": {
          "ignoreCase": false
        }
      },
      "type": "n8n-nodes-base.if",
      "typeVersion": 2.2,
      "position": [
        416,
        0
      ],
      "id": "99951517-a787-4b4e-b302-9259d586920b",
      "name": "If",
      "alwaysOutputData": false
    },
    {
      "parameters": {
        "message": "Thank you for using the Campaign Design chat window. I've gathered all the necessary details, and you'll receive a summary email shortly with everything we've captured. If there's anything else you'd like to add or clarify, feel free to let me know!",
        "waitUserReply": false,
        "options": {}
      },
      "type": "@n8n/n8n-nodes-langchain.chat",
      "typeVersion": 1,
      "position": [
        864,
        -96
      ],
      "id": "ac16e567-eeca-41ec-a31c-2a3dd891540a",
      "name": "Respond to Chat1"
    },
    {
      "parameters": {
        "message": "={{ $('AI Agent').item.json.output }}",
        "waitUserReply": false,
        "options": {}
      },
      "type": "@n8n/n8n-nodes-langchain.chat",
      "typeVersion": 1,
      "position": [
        640,
        96
      ],
      "id": "dd9357dc-5f42-4e09-8c15-2b15bdd5b221",
      "name": "Respond to Chat"
    },
    {
      "parameters": {
        "tableId": "campaign_capture_byNova",
        "fieldsUi": {
          "fieldValues": [
            {
              "fieldId": "campaign",
              "fieldValue": "={{ $('If').item.json.campaignPayload }}"
            },
            {
              "fieldId": "summaryText",
              "fieldValue": "={{ $('If').item.json.summaryText }}"
            },
            {
              "fieldId": "missingFields",
              "fieldValue": "={{ $('If').item.json.missingFields }}"
            }
          ]
        }
      },
      "type": "n8n-nodes-base.supabase",
      "typeVersion": 1,
      "position": [
        640,
        -96
      ],
      "id": "962f6da1-dd7b-4683-9e6d-7f539c02f5ed",
      "name": "Create a row",
      "credentials": {
        "supabaseApi": {
          "name": "<your credential>"
        }
      }
    }
  ],
  "connections": {
    "OpenAI Chat Model": {
      "ai_languageModel": [
        [
          {
            "node": "AI Agent",
            "type": "ai_languageModel",
            "index": 0
          }
        ]
      ]
    },
    "Simple Memory": {
      "ai_memory": [
        [
          {
            "node": "AI Agent",
            "type": "ai_memory",
            "index": 0
          }
        ]
      ]
    },
    "AI Agent": {
      "main": [
        [
          {
            "node": "Code in JavaScript",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "When chat message received": {
      "main": [
        [
          {
            "node": "AI Agent",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Code in JavaScript": {
      "main": [
        [
          {
            "node": "If",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "If": {
      "main": [
        [
          {
            "node": "Create a row",
            "type": "main",
            "index": 0
          }
        ],
        [
          {
            "node": "Respond to Chat",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Create a row": {
      "main": [
        [
          {
            "node": "Respond to Chat1",
            "type": "main",
            "index": 0
          }
        ]
      ]
    }
  },
  "active": false,
  "settings": {
    "executionOrder": "v1"
  },
  "versionId": "3b8761fe-437c-4c8f-883d-c58a843224c0",
  "meta": {
    "templateCredsSetupCompleted": true
  },
  "id": "YZ1uK5Rhb1YJtDEV",
  "tags": []
}

Credentials you'll need

Each integration node will prompt for credentials when you import. We strip credential IDs before publishing — you'll add your own.

• openAiApi
• supabaseApi

About this workflow

MarketingManager. Uses agent, lmChatOpenAi, memoryBufferWindow, chatTrigger. Chat trigger; 9 nodes.

Source: https://github.com/ideepak3m/MarketingManager/blob/216a213dd359a96e83038b4dc250eaff7cbba5ce/n8n/MarketingManager.json — original creator credit. Request a take-down →

More AI & RAG workflows → · Browse all categories →

Related workflows

Workflows that share integrations, category, or trigger type with this one. All free to copy and import.