Multi-Channel Customer Sentiment Tracker with Real-Time Analytics and Alerting

This n8n workflow automates the process of analyzing customer sentiment from various sources, recording it in a Google Sheet, storing raw data in a PostgreSQL database, and sending real-time alerts to Slack and email for critical sentiments. It helps businesses monitor customer feedback effectively and react promptly to both positive and negative trends.

What it does

1. Triggers on a Schedule: The workflow starts at predefined intervals (e.g., daily, hourly) to process new customer feedback.
2. Fetches Customer Feedback (Placeholder): An HTTP Request node is included as a placeholder for fetching customer feedback from various sources (e.g., social media APIs, CRM systems, survey platforms). You'll need to configure this node to connect to your specific data sources.
3. Analyzes Sentiment: The Sentiment Analysis node, powered by an AI language model (either OpenAI or Azure OpenAI Chat Model), processes the fetched text to determine the sentiment (e.g., positive, negative, neutral).
4. Categorizes Sentiment: An If node checks the sentiment score.
5. Records Sentiment in Google Sheets: Regardless of sentiment, each analyzed piece of feedback, along with its sentiment, is appended as a new row in a specified Google Sheet.
6. Stores Raw Data in PostgreSQL: The original feedback and its sentiment are also inserted into a PostgreSQL database for comprehensive historical tracking and advanced analytics.
7. Alerts for Critical Sentiments:
   • If the sentiment is identified as "negative" (or another critical threshold), the workflow sends a notification to a designated Slack channel.
   • Additionally, an email alert is dispatched to relevant stakeholders for immediate attention.
8. Aggregates Data (Placeholder): A Merge and Aggregate node are included, suggesting potential future enhancements for combining data from multiple sources or summarizing sentiment trends before further processing or reporting.
9. Transforms Data: Edit Fields (Set) and Code nodes are available for data manipulation, cleaning, or formatting as needed throughout the workflow.

Prerequisites/Requirements

• n8n Instance: A running n8n instance.
• OpenAI API Key or Azure OpenAI Chat Model Deployment: For the Sentiment Analysis node to function, you will need credentials for either OpenAI or an Azure OpenAI Chat Model.
• Google Sheets Account: A Google account with access to Google Sheets to store sentiment data.
• PostgreSQL Database: Access to a PostgreSQL database for storing raw feedback data.
• Slack Account: A Slack workspace and a channel to post alerts for critical sentiments.
• SMTP Email Server: Configuration for sending emails (e.g., SMTP server details, credentials).
• Customer Feedback Source: You will need to configure the HTTP Request node to connect to your specific customer feedback sources (e.g., Twitter API, Facebook API, Zendesk API, SurveyMonkey API, etc.).

Setup/Usage

1. Import the Workflow: Download the provided JSON and import it into your n8n instance.
2. Configure Credentials:
   • OpenAI/Azure OpenAI Chat Model: Set up your OpenAI or Azure OpenAI Chat Model credential in n8n.
   • Google Sheets: Authenticate your Google Sheets account.
   • PostgreSQL: Configure your PostgreSQL database credentials.
   • Slack: Authenticate your Slack workspace.
   • Send Email: Configure your SMTP email credentials.
3. Customize Schedule Trigger: Adjust the Schedule Trigger node to run at your desired frequency (e.g., every hour, once a day).
4. Configure HTTP Request:
   • Edit the HTTP Request node to connect to your specific customer feedback sources. You'll need to specify the API endpoint, authentication method, and any parameters required to fetch the feedback data.
   • Ensure the output format of this node is compatible with the Sentiment Analysis node (i.e., the text to be analyzed should be easily accessible in the item data).
5. Customize Sentiment Analysis:
   • Select the appropriate Language Model (OpenAI or Azure OpenAI Chat Model) and ensure it's configured with the correct credentials.
   • Map the input field to the customer feedback text from the previous node.
6. Configure If Node:
   • Adjust the conditions in the If node to define what constitutes a "critical" sentiment (e.g., sentiment.score < 0.3 for negative, or specific sentiment labels).
7. Configure Google Sheets Node:
   • Specify the Spreadsheet ID and Sheet Name where you want to record the sentiment data.
   • Map the fields to write (e.g., original feedback text, sentiment score, sentiment label, timestamp).
8. Configure Postgres Node:
   • Specify the table name in your PostgreSQL database.
   • Map the fields to insert (e.g., original feedback text, sentiment score, sentiment label, source, timestamp).
9. Configure Slack Node (for critical sentiments):
   • Specify the Slack Channel ID where alerts should be posted.
   • Customize the message to include relevant details about the critical feedback.
10. Configure Send Email Node (for critical sentiments):
    • Enter the recipient email address(es).
    • Customize the subject and body of the email alert.
11. Activate the Workflow: Once configured, activate the workflow to start automating your sentiment tracking.
12. (Optional) Customize Edit Fields (Set) and Code Nodes: Use these nodes to refine data before or after sentiment analysis, such as extracting specific information, cleaning text, or reformatting data.
13. (Optional) Implement Merge and Aggregate: If fetching from multiple sources or needing to summarize data, configure these nodes to combine and process the information as required.