calendar Last updated: 17 November 2026
YouTube and Claude logos, YouTube competitor analysis with Claude
Data integration
Data Pipelines
How to's
Youtube

YouTube Competitor Analysis with Claude: Track Any Public Channel

Try It Now

Connect your data in 1 min. Start free, no credit card

Got insights from this post? Give it a boost by sharing with others!

To run a YouTube competitor analysis in Claude, add the public channels you want to track to the YouTube Public connector in Windsor.ai, add Windsor.ai to Claude, and ask about those channels in a chat. You sign in with any Google account. You do not need access to the other channels, and nothing is scraped.

This guide covers channels you do not own: competitors, creators you might sponsor and the leaders in your niche. For your own channel’s analytics (watch time, retention, traffic sources, search terms), see how to connect YouTube Analytics to Claude.

A YouTube competitor analysis compares other channels in your niche with your own, to find the topics, formats and posting habits that get views. Claude ranks the videos, finds the ones that did far better than a channel’s usual result and groups titles and tags by theme.

We tested it on 17 November 2026 with windsor.ai’s channel and three channels in our niche: Supermetrics, Porter Metrics and Coupler.io Academy. Over the last 90 days, videos about connecting a tool to Claude were the outliers on three of the four channels.

🚀 Start tracking competitor channels with the free forever plan: https://onboard.windsor.ai/

What Claude can and cannot see on a public YouTube channel

The YouTube Public connector returns what YouTube shows any visitor, as tables Claude can compute on.

Claude can see:

  • Channel: title, @handle, country, creation date, subscriber count, public video count and total views.
  • Every public video: title, publish date, length, tags, thumbnail URL, views, likes and comments.
  • What Claude works out from those fields: views per video against the channel’s median, likes and comments per 1,000 views, uploads per week, the mix of video lengths, and recurring words in titles and tags.

Claude cannot see:

  • Watch time, retention, traffic sources, search terms, impressions, click-through rate or audience demographics. The YouTube Analytics API returns these only for a channel you are signed in to (Google’s API reference). No tool gets them for someone else’s channel.
  • Hidden numbers. When a channel hides its subscriber count or likes, the field is empty.
  • Exact subscriber counts. YouTube rounds public subscriber counts to three significant figures (channels resource): 1,234,567 subscribers shows as 1,230,000.
  • Which videos are Shorts. The connector has a length field and no Shorts flag, and Shorts can be up to three minutes long (YouTube Help).

Ways to get competitor YouTube data into Claude

These are the routes that put competitor YouTube data in front of Claude.

  • Windsor.ai YouTube Public connector: you add channels once and ask Claude questions across all of them, next to your own YouTube, Instagram or ad data from the same connector. It is read-only, and the free forever plan covers one account.
  • OutlierKit MCP server: a paid YouTube research tool with outlier scores, similar channels and sponsor data. You add its MCP server to Claude as a custom connector, and its pricing page lists MCP access on the Pro plan and above (OutlierKit’s Claude guide).
  • Metricool MCP: Metricool tracks competitor channels inside its social media suite and can pass that data to Claude through its MCP server (Metricool’s guide).
  • A Claude skill with your own API key: open-source skills such as youtube-competitor-analyzer call the YouTube Data API v3 from Claude Code. You need a Google Cloud project and an API key, and you manage the quota.
  • vidIQ or TubeBuddy: browser extensions that show competitor stats inside YouTube itself. Useful while you browse, but their numbers stay in their own interface.

Connect YouTube Public data to Claude in 3 steps

Setup takes a few minutes and needs no code. The YouTube Public to Claude integration page has the same steps in short form.

Prerequisites

  • A Windsor.ai account (the free forever plan is enough for a first test).
  • A Claude account on a plan that supports connectors.
  • The @handles of the channels you want to track, for example @porter.metrics.

Step 1: Add the channels in Windsor.ai

Log in to onboard.windsor.ai, select YouTube Public as the data source and click Sign in with Google. Any Google account works, because the connector reads public data only. Type a channel’s @handle in Channel to follow and click Add Account. Repeat for each channel. The channels appear under Account to follow, where Remove takes one off the list.

The YouTube Public form in Windsor.ai with four channels added: @windsorai, @coupleracademy, @porter.metrics and @supermetrics

Step 2: Add the Windsor.ai connector in Claude

In Claude, open Plugins in the sidebar, search for Windsor.ai and click Add. Sign in to your Windsor.ai account when Claude asks. If you have already added Windsor.ai for another source, skip this step: the same connector reaches every source in your account.

Step 3: Ask Claude about the channels

Start a new chat and name the channels in your question, for example: “Using YouTube Public in Windsor.ai, compare subscribers, video count and total views for the channels I track.”

Claude asks before each call to Windsor.ai. Allow it once, or always for this connector.

Claude's answer: the four tracked channels sorted by subscribers, with videos and total views

What our test showed, and what needed a second look

We added the four channels on 17 November 2026 and ran the snapshot and outlier prompts in Claude. Porter Metrics’ channel page showed 9.92K subscribers and 523 videos, the same as Claude’s table.

  • “Connect X to Claude” videos led on three channels. On Coupler.io Academy, Porter Metrics and windsor.ai, almost every outlier from the last 90 days was a Claude setup or Claude use-case video.
  • Coupler.io Academy posts mostly Shorts-length videos. 50 of its 71 videos in those 90 days were under one minute, and it had the highest median views of the four.
  • Bulk uploads pulled the median down. Supermetrics uploaded 19 API tutorials on 9 November, most with under 20 views. That put its median at 26, and videos with under 90 views counted as outliers. Our own channel had the same problem: 94 uploads, many of them ChatGPT versions of our Claude setup videos. Claude flagged both cases itself. When a channel uploads in batches, ask Claude to leave out same-day batches or to set a minimum view count before it ranks outliers.
  • Older videos have had more time to collect views. The view counts are lifetime totals: a video from August has had three months to collect views, one from last week a few days. Asking for views per day since publishing makes them comparable.
  • Claude guessed at a number it cannot see. In the snapshot answer it said Supermetrics’ videos “seem to pick up views from search or suggested videos”. The connector has no traffic sources for other channels. Treat any claim about watch time, traffic sources or audience in a competitor answer as a guess unless you check it.
  • The outlier question took about a minute. Claude pulled 272 videos and ran its own calculation to get each channel’s median.

Claude's outlier list for Coupler.io Academy: 71 videos, median 185 views, videos above 555 views

YouTube competitor analysis prompts for Claude

The prompts below use the YouTube Public connector’s field names.

Channel snapshot

Show channel_title, subscriber_count, video_count and view_count for every channel I track in YouTube Public, sorted by subscribers.

Outlier videos

For each channel, list the videos whose video_view_count is more than three times that channel's median, with video_title, published_at and video_length.

An outlier video tells you which topic or format found an audience beyond the channel’s subscribers. Use the median: one viral video can lift a channel’s average far above its usual result.

Upload cadence

Count each channel's videos by week of published_at for the last 90 days and show the average uploads per week.

The date range selects videos by their publish date. The free forever plan keeps 30 days of history. A 90-day question needs a paid plan. Our test account is on an Enterprise plan, and a two-year range returned Supermetrics videos back to November 2024.

Formats and length

Group each channel's videos into length bands (under 1 minute, 1 to 3 minutes, 3 to 10 minutes, over 10 minutes) from video_length and compare the median views per band.

Titles and tags

From video_title and video_tags, find the topics each competitor covers most often, and list topics that get high views on one channel and that the others never cover.

Engagement per video

Rank the last 20 videos from each channel by (video_like_count + video_comment_count) per 1,000 views.

Likes and comments are the only engagement numbers YouTube makes public for other channels.

Your channel against the competitors

Compare my channel's last 20 videos from the YouTube connector with the competitors' last 20 from YouTube Public: median views, uploads per week and most common length band.

This needs two data sources (YouTube for your channel, YouTube Public for the others). The free forever plan keeps one source after the first 30 days. Basic, at $19 a month for 3 sources, covers this question.

YouTube Public fields Claude can read through Windsor.ai

The YouTube Public connector has two groups of fields. The full list with descriptions is on the YouTube Public fields page.

  • Channel: channel_id, channel_title, channel_handle, channel_country, channel_created_date, subscriber_count, video_count, view_count.
  • Video: video, video_title, published_at, video_length, video_tags, video_thumbnail_url, video_view_count, video_like_count, video_comment_count.

The counts are current totals at the time of the query. A 7-day query in our test returned one row per channel, dated the last day of the range. YouTube Public keeps no subscriber history. To build one, you can schedule a daily export of subscriber_count to Google Sheets or BigQuery and let the rows add up.

What teams use YouTube competitor tracking for

  • Content planning: the outlier list from three or four competitors gives a list of topics with proven demand in your niche.
  • Agency pitches and reports: a client’s channel next to its three closest competitors, with the same metrics for each.
  • Creator vetting before a sponsorship: views per video and engagement per 1,000 views over the last 20 uploads say more than the subscriber count.
  • Cross-platform benchmarking: the same competitors’ Instagram accounts through the Instagram Public connector, in the same chat.

Conclusion

The YouTube Public connector gives Claude the public numbers of any channel: subscribers, videos, views, likes and comments. In one chat you can compare competitors, find their outlier videos and plan topics. It does not give watch time, retention or traffic sources for other channels, because YouTube keeps those for the channel owner.

Next steps:

FAQs

What is the difference between YouTube Public and YouTube in Windsor.ai?

YouTube Public reads public data from any channel: subscribers, videos, views, likes and comments. YouTube reads the private analytics of a channel you own or manage, such as watch time, retention, traffic sources and search terms. Use YouTube Public for competitors and YouTube for your own channel.

Tired of juggling fragmented data? Get started with Windsor.ai today to create a single source of truth

Let us help you automate data integration and AI-driven insights, so you can focus on what matters, growth strategy.
g logo
fb logo
big query data
youtube logo
power logo
looker logo