YouTube Comment Insights Workflow (n8n) + Gemini AI

Turn YouTube comments into real audience insights—automatically (n8n + Gemini AI)

Paste a public YouTube video URL into this n8n workflow and instantly generate an evidence-backed report of what viewers are repeatedly asking, struggling with, objecting to, and requesting—using Gemini AI to mine the actual comments.

What this workflow does

• Fetch public YouTube data: Uses the YouTube Data API v3 to retrieve the video’s title, channel, views, likes, comment count, and public comment threads.
• Select your analysis scope: Enter a public YouTube URL, choose how many comments to analyze, and pick relevance or time-based ordering.
• Clean and deduplicate comments: Comments and available replies are cleaned, spam-like/low-value comments are filtered, duplicate text is removed, and useful comments receive stable evidence IDs (e.g., C001).
• Mine recurring audience signals with Gemini: Gemini analyzes only the collected comments to identify recurring questions, problems, complaints, feature requests, objections, tutorial requests, purchase intent, and other audience signals.
• Convert signals into action: The workflow turns the strongest repeated patterns into content ideas, hooks, and product/service opportunities—while keeping one-off observations separate.
• Generate an n8n report with clickable evidence: The final output is rendered directly in the n8n form, including an executive summary, audience signals, engagement context, opportunities, caveats, and links back to the original YouTube comments.

Use cases

• Find recurring viewer questions to shape your next YouTube tutorial, blog post, or SaaS onboarding content.
• Spot objections and feature requests to prioritize product roadmap decisions.
• Generate market research from comments for a new offer, landing page angle, or webinar topic.

Technical details

• Google Cloud setup: create a project and enable YouTube Data API v3, then add an API key.
• n8n nodes/workflow blocks include: Code, Form Trigger, HTTP Request, Form, Sticky Note, and n8nn8n-nodes-langchainagent.