YouTube is the world's second largest search engine. Researching trends, analyzing competitors, and managing your own channel manually takes hours that could be automated.
Companion connects to YouTube and handles the research layer. Here's how.
What you need
- Companion installed (download here)
- An OpenRouter API key (free, pay per use)
- A Google account (for YouTube channel management features)
Step 1: Connect YouTube
Open Settings → Integrations in Companion's sidebar. Find YouTube and click Connect.
Authorize via Google OAuth — the same account linked to your YouTube channel. Companion requests access to read channel data and analytics, and optionally to manage uploads and comments.
Step 2: Your first automation
Start with competitive research:
Search YouTube for the top 10 videos about "MCP AI agents" from the last 3 months. List the titles, channels, view counts, and a one-sentence summary of each.
Companion searches YouTube, reads the results, and gives you a formatted research summary — without you opening YouTube, searching, clicking through videos, and taking notes manually.
Or for channel management:
Check my YouTube channel analytics from the last 7 days: views, subscribers gained, top performing video, and any comments that need a reply.
Companion pulls your analytics and surfaces what matters, including flagging comments that are questions or feedback worth responding to.
What it costs
YouTube API calls and browser-based research use roughly 2,000–5,000 tokens per research session. At Claude Haiku rates, a thorough competitive research session costs under $0.005. Compare that to 30–60 minutes of manual research time.
More workflows
Content ideation: "What are the most searched YouTube topics about AI desktop tools right now? What questions are people asking that don't have great video answers yet?"
Comment moderation: "Check my YouTube comments from the last 48 hours. List any questions or negative comments that I should respond to, with the commenter's name and what they said."
Video description writing: "Write a YouTube description for a video called 'How to Install MCP in Companion.' Include keywords, a brief summary, timestamps, and links to the download page."
Competitor analysis: "Find the top 5 YouTube channels covering AI productivity tools. How often do they post, what gets the most views, and what topics do they cover most?"
Thumbnail research: "Search YouTube for 'AI agent desktop' and describe the thumbnail styles used in the top 10 results — what colors, text, faces, and layouts are most common?"
Combine with Notion: "Research YouTube trends for AI tools this week and add a summary to my 'Content Ideas' Notion database with topic, potential title, and why it's timely"
For content creators and marketers, combining YouTube research with blog drafting means Companion can go from "here's what's trending" to a draft blog post or script in the same session.
Achmad Bifari is Creative Designer and Social Manager at Companion. He writes the tutorials and runs @companionbyaios.