Track Weekly ChatGPT Citations with n8n Data Tables

Track weekly whether ChatGPT cites your site—automatically with n8n

This n8n workflow sends your buyer questions to the OpenAI Responses API with web search enabled every week, extracts all returned citation URLs, checks if your domain is cited and your brand is mentioned, and stores a weekly scorecard in an n8n Data Table called ai_citation_results.

What this workflow does

• Schedules weekly (or runs manually): It runs every Monday at 8:00 AM or can be triggered on demand.
• Loads your settings: Reads your configured brand domain, brand name, model, and an array of buyer questions.
• Queries OpenAI with web search: For each question, it calls https://api.openai.com/v1/responses and uses the web_search tool.
• Extracts citations and context flags: It captures the answer text, whether web search occurred, and all URL citations returned by OpenAI.
• Scores your presence: It checks whether any cited URL contains your domain, whether the answer mentions your brand name, counts the number of sources, and records the cited URLs.
• Saves a weekly row: Inserts run_date, prompt, cited, mentioned, sources, cited_urls into the n8n Data Table.

Use cases

• Monitor how often ChatGPT cites your site for specific buyer questions over time.
• Detect brand visibility changes (your domain cited vs. brand name mentioned) after website updates.
• Compare outcomes across multiple models by swapping the model in workflow Settings.

Technical details

• OpenAI integration: OpenAI Responses API (HTTP request to /v1/responses) with web_search enabled.
• n8n Data Tables: Stores results in ai_citation_results (columns: run_date, prompt, cited, mentioned, sources, cited_urls).
• Workflow nodes used: set, split out, data table, sticky note, http request, manual trigger.

Setup checklist: Add an OpenAI API credential for the HTTP request, create the Data Table named ai_citation_results, and update Settings for brand_domain, brand_name, model, and your prompts.