klbr/n8nworkflows.xyz
1
0
Fork 0
mirror of https://github.com/khoaliber/n8nworkflows.xyz.git synced 2026-04-28 00:29:22 +00:00
Code Issues Packages Projects Releases Wiki Activity

creation

Browse Source
This commit is contained in:
nusquama
2025-11-12 16:07:06 +01:00
parent 5d1fa08532
commit 5b94fdccd4
1 changed files with 409 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
workflows/Optimize Amazon Ads with GPT-4o for Bid, Budget & Keyword Recommendations-3793/readme-3793.md
+409
View File
@@ -0,0 +1,409 @@
Optimize Amazon Ads with GPT-4o for Bid, Budget & Keyword Recommendations
https://n8nworkflows.xyz/workflows/optimize-amazon-ads-with-gpt-4o-for-bid--budget---keyword-recommendations-3793
# Optimize Amazon Ads with GPT-4o for Bid, Budget & Keyword Recommendations
### 1. Workflow Overview
This workflow automates the optimization of Amazon Sponsored Products advertising campaigns by analyzing performance reports and generating actionable recommendations using OpenAI’s GPT-4o model. It targets Amazon sellers and advertisers seeking to streamline campaign analysis, bidding, budgeting, and keyword management without manual report review.
The workflow is logically divided into these blocks:
- **1.1 Input Reception & File Acquisition**
Automatically lists and downloads Amazon Ads reports stored in a specified Google Drive folder.
- **1.2 File Type Detection & Data Extraction**
Differentiates between XLSX and CSV report files, extracting structured data accordingly.
- **1.3 Data Normalization & Aggregation**
Normalizes filenames, classifies reports into categories (search terms, campaigns, targeting, placement, budgets), and merges data for AI processing.
- **1.4 AI Analysis & Recommendation Generation**
Sends cleaned, structured data to GPT-4o with detailed instructions to produce bid, budget, and keyword recommendations.
- **1.5 Email Notification**
Formats the AI output into a human-readable email and sends it via Gmail to configured recipients.
- **1.6 Workflow Trigger & Configuration**
Manual trigger and email option setup nodes facilitate testing and configuration.
---
### 2. Block-by-Block Analysis
#### 2.1 Input Reception & File Acquisition
**Overview:**
This block lists files from a user-specified Google Drive folder containing Amazon Ads reports and downloads each file for processing.
**Nodes Involved:**
- List Files
- Get File
- Set fileName
**Node Details:**
- **List Files**
- Type: Google Drive (fileFolder resource, query method)
- Role: Lists files in a specified Drive folder (configured via folderId filter)
- Key Config: Folder ID must be set to the folder containing Amazon Ads reports
- Input: Trigger from "Email Options" node
- Output: List of files metadata (id, name)
- Failure Modes: Invalid folder ID, permission errors, API rate limits
- **Get File**
- Type: Google Drive (download operation)
- Role: Downloads each listed file’s binary content for further processing
- Key Config: Uses fileId from previous node, downloads file as binary property "data"
- Input: Files from "List Files"
- Output: Binary file content with metadata
- Failure Modes: File not found, download errors, permission issues
- **Set fileName**
- Type: Set
- Role: Extracts and assigns the filename from the downloaded file metadata to a JSON property `fileName` for downstream use
- Input: Downloaded file data from "Get File"
- Output: Adds `fileName` property to JSON
- Failure Modes: Missing or malformed filename
---
#### 2.2 File Type Detection & Data Extraction
**Overview:**
Determines whether each file is XLSX or CSV and extracts data accordingly.
**Nodes Involved:**
- is XLSX (If node)
- Extract XLSX Data
- Extract CSV Data
- Preserve File Name (for XLSX)
- Preserve CSV File Name
**Node Details:**
- **is XLSX**
- Type: If
- Role: Checks if filename contains ".xlsx" (case-sensitive) to branch extraction logic
- Input: JSON with `fileName`
- Output: Two branches: XLSX or CSV
- Failure Modes: Filename missing or unexpected format
- **Extract XLSX Data**
- Type: Extract From File (xlsx operation)
- Role: Parses XLSX binary data into JSON rows
- Input: Binary data from "Get File" when XLSX branch is true
- Output: JSON array of rows from the spreadsheet
- Failure Modes: Corrupt XLSX, unsupported formats
- **Extract CSV Data**
- Type: Extract From File (default CSV)
- Role: Parses CSV binary data into JSON rows
- Input: Binary data from "Get File" when XLSX branch is false
- Output: JSON array of rows from CSV
- Failure Modes: Malformed CSV, encoding issues
- **Preserve File Name (XLSX & CSV)**
- Type: Set
- Role: Copies the filename from the XLSX branch to CSV branch or vice versa to maintain consistent metadata for downstream merging
- Input: Extracted data from respective extraction nodes
- Output: Adds `fileName` property to JSON
- Failure Modes: Missing filename propagation
---
#### 2.3 Data Normalization & Aggregation
**Overview:**
Merges extracted data from both XLSX and CSV files and normalizes filenames to classify reports into predefined categories for AI input.
**Nodes Involved:**
- Merge XLSX and CSV
- Format Data
**Node Details:**
- **Merge XLSX and CSV**
- Type: Merge
- Role: Combines data arrays from XLSX and CSV extraction paths into a single stream
- Input: Outputs from "Preserve File Name" nodes for XLSX and CSV
- Output: Combined dataset for processing
- Failure Modes: Data format mismatches
- **Format Data**
- Type: Code
- Role: Normalizes filenames by removing spaces, suffixes, and converting to lowercase; maps files to keys: search_terms, campaigns, targeting, placement, budgets
- Key Logic: Uses regex matching on normalized base filename to assign category keys
- Input: Merged data with filenames
- Output: JSON object with arrays keyed by report type
- Failure Modes: Unrecognized filenames cause workflow errors; malformed data
---
#### 2.4 AI Analysis & Recommendation Generation
**Overview:**
Sends the structured input data to an AI model (GPT-4o) with detailed instructions to generate bid, budget, keyword, and targeting recommendations.
**Nodes Involved:**
- AI Analyze (Chain LLM)
- OpenAI Chat Model (GPT-4o)
**Node Details:**
- **AI Analyze**
- Type: Chain LLM (LangChain)
- Role: Wraps the OpenAI Chat Model node, sends JSON stringified structured data with system prompt defining expected JSON output format
- Key Config: System prompt instructs AI to return campaign adjustments, keyword recommendations, and targeting recommendations in strict JSON format without explanations
- Input: JSON from "Format Data"
- Output: AI-generated JSON recommendations
- Failure Modes: API errors, malformed AI output, rate limits, timeout, JSON parse errors downstream
- **OpenAI Chat Model**
- Type: OpenAI Chat Completion
- Role: Executes GPT-4o model call with prompt from "AI Analyze"
- Input: Prompt and messages from "AI Analyze"
- Output: Raw AI response text
- Failure Modes: Authentication errors, quota exceeded, network issues
---
#### 2.5 Email Notification
**Overview:**
Formats AI-generated JSON recommendations into a structured HTML email and sends it to configured recipients via Gmail.
**Nodes Involved:**
- Email Optimizations
- Email Options
**Node Details:**
- **Email Options**
- Type: Set
- Role: Defines email recipient address and subject line for outgoing optimization emails
- Input: Manual trigger or workflow start
- Output: JSON with `send_to` and `subject` fields
- Failure Modes: Missing or invalid email addresses
- **Email Optimizations**
- Type: Gmail (OAuth2)
- Role: Parses AI JSON output, formats it into a detailed HTML email including campaign adjustments, keyword and targeting recommendations, and sends it via Gmail
- Key Logic:
- Removes markdown code blocks from AI output
- Parses JSON safely with error handling
- Summarizes total budget increase, estimated spend and sales
- Lists campaign bid and budget changes, keyword additions and negatives, targeting actions
- Input: AI Analyze output and Email Options
- Output: Sent email confirmation
- Failure Modes: Email sending errors, invalid JSON, missing credentials
---
#### 2.6 Workflow Trigger & Configuration
**Overview:**
Facilitates manual testing and initial configuration of email parameters.
**Nodes Involved:**
- When clicking ‘Test workflow’ (Manual Trigger)
- Sticky Notes (multiple)
**Node Details:**
- **When clicking ‘Test workflow’**
- Type: Manual Trigger
- Role: Starts the workflow manually for testing
- Output: Triggers "Email Options" node to start the chain
- Failure Modes: None
- **Sticky Notes**
- Type: Sticky Note
- Role: Provide detailed instructions and guidance for users on folder selection, report scheduling, report delivery methods, upgrading to API automation, and AI analysis explanation
- Content Highlights:
- Folder selection for reports
- Amazon Ads report scheduling instructions (report types, frequency, format)
- Report delivery options (manual upload, Gmail automation, Drive sync)
- Upgrade path to Amazon Ads API for full automation
- AI analysis explanation and expected output format
- Failure Modes: None (informational only)
---
### 3. Summary Table
| Node Name | Node Type | Functional Role | Input Node(s) | Output Node(s) | Sticky Note |
|-------------------------|--------------------------------|-----------------------------------------------|-------------------------|--------------------------|----------------------------------------------------------------------------------------------|
| When clicking ‘Test workflow’ | Manual Trigger | Manual workflow start trigger | | Email Options | ## Trigger You may replace this with a scheduled event or poll the folder for changes. |
| Email Options | Set | Configures email recipient and subject | When clicking ‘Test workflow’ | List Files | ## Change! Edit these email options. |
| List Files | Google Drive | Lists files in Amazon Ads reports folder | Email Options | Get File | ## Change Choose the "folder" in the filter options to the folder containing your Ad reports |
| Get File | Google Drive | Downloads each file listed | List Files | Set fileName | |
| Set fileName | Set | Assigns filename property for downstream use | Get File | is XLSX | |
| is XLSX | If | Checks if file is XLSX or CSV | Set fileName | Extract XLSX Data, Extract CSV Data | |
| Extract XLSX Data | Extract From File (XLSX) | Extracts data from XLSX files | is XLSX (true) | Preserve File Name | |
| Extract CSV Data | Extract From File (CSV) | Extracts data from CSV files | is XLSX (false) | Preserve CSV File Name | |
| Preserve File Name | Set | Preserves filename metadata for XLSX branch | Extract XLSX Data | Merge XLSX and CSV | |
| Preserve CSV File Name | Set | Preserves filename metadata for CSV branch | Extract CSV Data | Merge XLSX and CSV | |
| Merge XLSX and CSV | Merge | Combines XLSX and CSV extracted data | Preserve File Name, Preserve CSV File Name | Format Data | |
| Format Data | Code | Normalizes filenames and maps reports to keys | Merge XLSX and CSV | AI Analyze | |
| AI Analyze | Chain LLM (LangChain) | Sends structured data to GPT-4o for analysis | Format Data | Email Optimizations | ## AI Analysis Uses GPT-4o to process the bundled reports and generate optimization instructions. Passes system instructions and cleaned data as input. |
| OpenAI Chat Model | OpenAI Chat Completion | Executes GPT-4o model call | AI Analyze (ai_languageModel) | AI Analyze | |
| Email Optimizations | Gmail (OAuth2) | Formats AI output and sends optimization email | AI Analyze, Email Options | | |
| Sticky Note | Sticky Note | Instructions: Trigger info | | | ## Trigger You may replace this with a scheduled event or poll the folder for changes. |
| Sticky Note1 | Sticky Note | Instructions: Email options | | | ## Change! Edit these email options. |
| Sticky Note2 | Sticky Note | Instructions: Folder selection | | | ## Change Choose the "folder" in the filter options to the folder containing your Ad reports |
| Sticky Note3 | Sticky Note | Instructions: AI analysis explanation | | | ## AI Analysis Uses GPT-4o to process the bundled reports and generate optimization instructions. Passes system instructions and cleaned data as input. |
| Sticky Note5 | Sticky Note | Instructions: Amazon Ads report scheduling | | | ## Amazon Ads Report Scheduling Instructions To run this workflow, schedule the following Sponsored Products reports in the Amazon Ads Console: Use "Detailed" for Search Term and Targeting reports, "Summary" for Campaign, Placement, and Budget reports. Use Last 30 Days, Daily frequency, and .xlsx or .csv format. |
| Sticky Note6 | Sticky Note | Instructions: Report delivery options | | | ## Report Delivery How to get reports into Google Drive: manual upload, Gmail automation, or Drive sync folder. Reports must match expected filenames. |
| Sticky Note7 | Sticky Note | Instructions: Upgrade to Amazon Ads API | | | ## Upgrade! 🚀 Apply for an Amazon Advertising API developer account to unlock full automation: https://advertising.amazon.com/API/docs/en-us/ |
---
### 4. Reproducing the Workflow from Scratch
1. **Create Manual Trigger Node**
- Name: When clicking ‘Test workflow’
- Type: Manual Trigger
- Purpose: To manually start the workflow for testing.
2. **Create Set Node for Email Options**
- Name: Email Options
- Type: Set
- Parameters:
- `send_to`: Set to recipient email address (e.g., your email)
- `subject`: Set to desired email subject for optimization reports
- Connect "When clicking ‘Test workflow’" → "Email Options"
3. **Create Google Drive List Files Node**
- Name: List Files
- Type: Google Drive (fileFolder resource, query method)
- Credentials: Connect your Google Drive OAuth2 credentials
- Parameters:
- Filter: Set folderId to the Google Drive folder containing Amazon Ads reports
- Connect "Email Options" → "List Files"
4. **Create Google Drive Get File Node**
- Name: Get File
- Type: Google Drive (download operation)
- Credentials: Same Google Drive OAuth2 credentials
- Parameters:
- fileId: Set to `={{ $json.id }}` from "List Files" output
- Binary Property Name: `data`
- Connect "List Files" → "Get File"
5. **Create Set Node to Assign Filename**
- Name: Set fileName
- Type: Set
- Parameters:
- Assign `fileName` to `={{ $json.name }}`
- Include other fields
- Connect "Get File" → "Set fileName"
6. **Create If Node to Check XLSX**
- Name: is XLSX
- Type: If
- Parameters:
- Condition: Check if `fileName` contains `.xlsx` (case-sensitive)
- Connect "Set fileName" → "is XLSX"
7. **Create Extract From File Node for XLSX**
- Name: Extract XLSX Data
- Type: Extract From File
- Parameters:
- Operation: `xlsx`
- Connect "is XLSX" (true branch) → "Extract XLSX Data"
8. **Create Extract From File Node for CSV**
- Name: Extract CSV Data
- Type: Extract From File
- Parameters:
- Binary Property Name: `data`
- Connect "is XLSX" (false branch) → "Extract CSV Data"
9. **Create Set Node to Preserve Filename for XLSX**
- Name: Preserve File Name
- Type: Set
- Parameters:
- Assign `fileName` from `={{ $('is XLSX').item.json.fileName }}`
- Include other fields
- Connect "Extract XLSX Data" → "Preserve File Name"
10. **Create Set Node to Preserve Filename for CSV**
- Name: Preserve CSV File Name
- Type: Set
- Parameters:
- Assign `fileName` from `={{ $('is XLSX').item.json.fileName }}`
- Include other fields
- Connect "Extract CSV Data" → "Preserve CSV File Name"
11. **Create Merge Node to Combine XLSX and CSV Data**
- Name: Merge XLSX and CSV
- Type: Merge
- Parameters: Default (merge by append)
- Connect "Preserve File Name" → "Merge XLSX and CSV" (input 1)
- Connect "Preserve CSV File Name" → "Merge XLSX and CSV" (input 2)
12. **Create Code Node to Format Data**
- Name: Format Data
- Type: Code
- Parameters: Use the provided JavaScript code to:
- Normalize filenames (remove spaces, suffixes, lowercase)
- Map files to keys: search_terms, campaigns, targeting, placement, budgets
- Throw error if filename is unrecognized
- Connect "Merge XLSX and CSV" → "Format Data"
13. **Create Chain LLM Node for AI Analysis**
- Name: AI Analyze
- Type: Chain LLM (LangChain)
- Credentials: OpenAI API key with GPT-4o access
- Parameters:
- Text: `={{JSON.stringify($json)}}`
- Prompt: Provide system instructions specifying expected JSON output format including campaign adjustments, keyword and targeting recommendations
- Connect "Format Data" → "AI Analyze"
14. **Create OpenAI Chat Model Node**
- Name: OpenAI Chat Model
- Type: OpenAI Chat Completion
- Credentials: Same OpenAI API key
- Parameters:
- Model: gpt-4o
- Connect "AI Analyze" (ai_languageModel output) → "OpenAI Chat Model"
15. **Create Gmail Node to Send Email**
- Name: Email Optimizations
- Type: Gmail (OAuth2)
- Credentials: Gmail OAuth2 credentials
- Parameters:
- Send To: `={{ $('Email Options').first().json.send_to }}`
- Subject: `={{ $('Email Options').first().json.subject }}`
- Message: JavaScript expression that:
- Parses AI JSON output safely
- Formats HTML email with campaign adjustments, keyword recommendations, targeting recommendations, and summary totals
- Connect "AI Analyze" → "Email Optimizations"
16. **Add Sticky Notes**
- Add sticky notes at appropriate places with instructions on:
- Folder selection for Google Drive
- Amazon Ads report scheduling details
- Report delivery options
- AI analysis explanation
- Upgrade path to Amazon Ads API automation
---
### 5. General Notes & Resources
| Note Content | Context or Link |
|------------------------------------------------------------------------------------------------------------------------------------|-------------------------------------------------------------------------------------------------|
| To run this workflow, schedule Sponsored Products reports in Amazon Ads Console with specific report types, date ranges, and formats. | Amazon Ads Report Scheduling Instructions sticky note content |
| Reports must be delivered to a Google Drive folder monitored by this workflow, either manually or via Gmail automation. | Report Delivery sticky note content |
| Upgrade to Amazon Advertising API developer account for full automation of report generation and retrieval. | Upgrade sticky note with link: https://advertising.amazon.com/API/docs/en-us/ |
| AI analysis uses GPT-4o to generate structured JSON recommendations for bids, budgets, keywords, and targeting adjustments. | AI Analysis sticky note content |
| Workflow trigger can be replaced with scheduled trigger or folder polling for automation. | Trigger sticky note content |
| Detailed instructions and field-specific guidance are embedded in sticky notes within the workflow for user reference. | Workflow description and sticky notes |
---
This documentation enables advanced users and AI agents to understand, reproduce, and extend the Amazon Ads AI Optimization workflow effectively, anticipating potential errors and integration points.