AI-Powered Umami Analytics Reporter & SEO Advisor

Unlock Actionable SEO Insights and Automate Website Performance Reporting

This AI Agent streamlines your website analytics workflow by connecting directly to your Umami instance. It fetches key performance indicators, including overall site statistics and page-specific data for the current and previous week. This data is then processed and sent to a powerful Large Language Model (LLM) like Llama 3.1 via OpenRouter, which acts as an SEO expert to analyze trends, summarize performance in markdown tables, and provide actionable improvement suggestions. Finally, these valuable insights are neatly organized and saved into your Baserow database, creating a historical record for ongoing strategy refinement.

Key Features & Benefits

• Automated Data Collection: Regularly fetches vital statistics (pageviews, visitors, visits, bounces, total time) and detailed page metrics from your Umami analytics.
• AI-Driven Analysis: Leverages an LLM (configurable, e.g., Llama 3.1 via OpenRouter) to interpret complex data, identify top-performing content, and pinpoint SEO opportunities.
• Comparative Reporting: Automatically generates reports comparing website performance week-over-week, highlighting changes and trends.
• Actionable SEO Recommendations: Receives concrete suggestions from the AI to improve content, optimize for search engines, and enhance user engagement.
• Structured Data Storage: Saves all generated reports and AI analyses into a Baserow table for easy access, historical tracking, and strategic decision-making.
• Flexible Scheduling: Can be scheduled to run automatically (e.g., weekly) or triggered manually for on-demand reporting.
• Customizable Prompts: Easily tailor the AI prompts to focus the analysis on specific aspects of your website or business goals.

• An n8n instance (Cloud or self-hosted).
• Umami instance URL and API access (Header Auth credentials specific to your Umami setup).
• OpenRouter API Key (the workflow is pre-configured for OpenRouter but can be adapted for other LLM providers via HTTP Request node).
• Baserow instance, API token, Database ID, and a pre-configured Table with the following fields (names can be adjusted, but types should match):
  • Date (date type)
  • Summary (long text type for overall AI summary)
  • Top pages (long text type for page-specific AI analysis)
  • Blog name (text type, or customize as needed).

1. Download the n8n workflow JSON file.
2. Import the workflow into your n8n instance.
3. Configure Umami Nodes:
   • For 'Get view stats from Umami', 'Get page data from Umami', and 'Get page view data from Umami' nodes: Update the URL parameter with your Umami instance address and specific Website ID. Adjust date parameters (startAt, endAt, unit, timezone) as needed.
   • Set up your Umami Header Auth credentials in each of these nodes.
4. Configure AI (OpenRouter) Nodes:
   • For 'Send data to A.I.' and 'Send data to A.I.1' (HTTP Request) nodes:
     • Ensure the URL is https://openrouter.ai/api/v1/chat/completions.
     • Set up Header Auth credentials for OpenRouter: Username Authorization, Password Bearer YOUR_OPENROUTER_API_KEY.
     • In the JSON Body, you can change the model (default meta-llama/llama-3.1-70b-instruct:free).
     • Customize the content (prompt) to reflect your website's name (e.g., replace 'Pennibnotes.com') and specific analytical needs.
5. Configure Baserow Node:
   • In the 'Save data to Baserow' node:
     • Select or create your Baserow API credentials.
     • Enter your Baserow Database ID and Table ID.
     • Map the Field Values to your Baserow table columns. Ensure expressions correctly reference data from previous AI nodes (e.g., {{ $('Send data to A.I.').first().json.choices[0].message.content }} for the summary).
6. Review Code Nodes:
   • The 'Parse Umami data', 'Parse Umami data1', and 'Parse Umami' Code nodes transform data from Umami before sending it to the AI. Review their JS code and adjust if your Umami API response or data requirements differ.
7. Set Trigger:
   • Configure the 'Schedule Trigger' node to your desired reporting frequency (default is weekly on Thursdays at midnight).
   • Alternatively, use the 'When clicking ‘Test workflow’' node for manual execution.
8. Activate the workflow.