AI-Powered Google Analytics SEO Reporter & Advisor

Unlock Actionable SEO Insights with this AI Agent

This AI Agent automates the weekly analysis of your Google Analytics 4 data, aiming to transform raw numbers into strategic SEO advice. It fetches key metrics such as page engagement, Google Search Console performance (clicks, impressions, CTR, average position), and country-specific views for both the current and previous week. This data is then processed. The workflow sends these datasets to a powerful Large Language Model (LLM) like Llama 3.1 via OpenRouter, with prompts designed to compare weekly data, generate performance summaries in markdown, and provide actionable SEO suggestions. Finally, these AI-generated insights are systematically saved to a Baserow table for easy tracking and historical review.

Key Features & Benefits

• Automated Data Collection: Connects directly to Google Analytics 4 to pull vital SEO metrics weekly, saving you hours of manual work.
• Comprehensive Metrics Retrieval: Gathers data on page views, active users, engagement rates, event counts, organic search performance (clicks, impressions, CTR, average position), and user data by country.
• AI-Powered SEO Analysis: Leverages advanced LLMs (e.g., Llama 3.1 via OpenRouter) to interpret data, with the goal of identifying trends and comparing week-over-week performance for page engagement, search, and country metrics.
• Actionable Recommendations: Delivers AI-generated SEO suggestions tailored to the analyzed data, helping you improve search rankings and user engagement.
• Structured Reporting: Outputs analysis in clear markdown tables and stores all insights neatly in Baserow, creating a historical record of your SEO efforts and AI advice.
• Customizable & Extensible: Easily adjust the reporting schedule, AI prompts, and metrics to fit your specific business needs. Can be triggered manually or run on a schedule.
• Time-Saving Automation: Frees up your team from tedious data gathering and initial analysis, allowing focus on strategic implementation.

• An n8n instance (Cloud or self-hosted).
• Google Analytics 4 Property ID and configured Google OAuth2 credentials in n8n.
• OpenRouter API Key (the workflow is pre-configured for OpenRouter and Llama 3.1, but can be adapted for other LLM providers).
• Baserow account, Database ID, Table ID, and API Token.
• A Baserow table pre-configured with fields to store the AI's analysis (e.g., Blog Name, Page Views Analysis, Search Report Analysis, Country Views Analysis, Report Date).

1. Download the n8n workflow JSON file.
2. Import the workflow into your n8n instance.
3. Google Analytics Nodes Configuration:
   • For each of the six Google Analytics nodes (e.g., 'Get Page Engagement Stats for this week', 'Get Google Search Results for this week'), select your pre-configured Google Analytics OAuth2 credentials.
   • In the parameters of each Google Analytics node, replace the placeholder 'Property ID' (e.g., 460520224) with your actual GA4 Property ID.
   • Review date configurations in Google Analytics nodes. For accurate week-over-week comparison, ensure 'prior week' nodes (e.g., 'Get Page Engagement Stats for prior week') use dynamic end dates like ={{$today.minus({days: 7})}} instead of fixed dates.
4. AI Request Nodes Configuration (Nodes: 'Send page data to A.I.', 'Send page Search data to A.I.', 'Send country view data to A.I.'):
   • Select 'Generic Credential Type' for Authentication and choose 'Header Auth'.
   • Create or select credentials: For Header Auth, use Authorization as the 'Name' and Bearer YOUR_OPENROUTER_API_KEY as the 'Value'. Replace YOUR_OPENROUTER_API_KEY with your actual OpenRouter API key (ensure there's a space after 'Bearer').
   • The model is set to meta-llama/llama-3.1-70b-instruct:free. You can change this if you prefer another model available on OpenRouter.
   • Review and customize the prompts within the content field of the JSON body if needed.
   • Crucial: Carefully verify that the n8n expressions used for 'Data from 2 weeks ago' (representing last week's data) and 'Data from last week' (representing this week's data) in each of the three 'Send ... to A.I.' HTTP Request nodes correctly reference the corresponding parsed datasets. For instance, the 'Send page data to A.I.' node should ideally use expressions pointing to last week's page engagement data and this week's page engagement data respectively. The provided template may require adjustments to these expressions for accurate comparisons.
5. Baserow Node Configuration ('Save A.I. output to Baserow'):
   • Select or create your Baserow API credentials.
   • Enter your Baserow 'Database ID' and 'Table ID'.
   • Ensure your Baserow table has fields corresponding to the data being mapped. The workflow maps to fields for: blog name, page view analysis, search report analysis, country view analysis, and report date. Adjust field mappings and the 'Blog Name' (fieldId: 5833, pre-filled as 'Name of your blog') as needed.
6. Review Code Nodes: The Code nodes ('Parse data from Google Analytics', 'Parse GA data', etc.) are designed to structure the GA data for the AI. Generally, no changes are needed unless you significantly alter the metrics fetched from Google Analytics or the data structure changes.
7. Set Schedule: The 'Schedule Trigger' node is set for weekly execution. Adjust the interval as required.
8. Test Workflow: Use the 'When clicking ‘Test workflow’' manual trigger or execute individual nodes to ensure correct configuration and data flow.
9. Activate the workflow to enable scheduled execution.