How to Connect YouTube Analytics to Claude: Automated AI Insights in 60 Seconds

You might know which videos went viral or flopped. But what you might not know is why it happened or what to do next.

YouTube Studio gives you basic charts, but it doesn’t tell you, for example, that your how-to format consistently outperforms your opinion pieces, or that watch time drops sharply at the three-minute mark across your last twenty uploads. That kind of insight usually requires an analyst or a lot of time that you’d be better off spending creating content.

Here’s a fully automated solution to save you time and effort: with the Windsor MCP connector, your YouTube Analytics data flows into Claude in just a few clicks.

🚀 Automate your YouTube reporting in Claude with a 30-day free trial: https://onboard.windsor.ai/.

Ask questions in plain English and get in-depth AI insights and visual summaries—like the ones shown below—in less than a minute. No more manual exports, dashboards, or analyst bottlenecks.

It only involves three simple steps.

3 steps to connect YouTube Analytics to Claude via Windsor MCP

Setting up the YouTube-to-Claude integration is a no-code process that takes less than 60 seconds.

📖 Full step-by-step guide: https://windsor.ai/documentation/windsor-mcp/how-to-integrate-data-into-claude/.

Prerequisites

• A Windsor.ai account (free trial or paid plan)
• A YouTube channel with YouTube Analytics access
• A Claude.ai account

Step 1: Connect your YouTube account in Windsor

1. Log in to onboard.windsor.ai.
2. Select YouTube as your data source
3. Authenticate with your Google account and grant Windsor access to your YouTube Analytics.
4. Select the YouTube channels that you’d like to pull data from.

Windsor connects to the YouTube Analytics API and automatically normalizes your channel and video data into a clean, structured format, ready for Claude to reason about.

Step 2: Add the Windsor.ai connector in Claude

Open the Windsor.ai connector page in Claude and add it with the Connect button.

💡 Quick tip: Set permissions to “Always allow” so your Windsor data is accessible across all chats without re-authorizing.

Step 3: Start asking Claude about your YouTube channel

No more switching between tools and tabs. Just ask simple questions like:

• “Which of my videos had the highest watch time percentage last month?”
• “Compare my Shorts performance vs. long-form videos over the last 90 days.”
• “Which traffic source is driving the most subscribers right now?”
• “Show me videos where click-through rate is high but average view duration is low — what might explain the gap?”

💡 Pro tip: Ask Claude to generate a weekly performance summary or suggest your next video topic based on what’s working — it can do both in a single prompt.

Bonus: YouTube Analytics prompts for Claude

These prompts are built around the specific metrics Windsor streams from YouTube Analytics, designed to surface the strategic insights that YouTube Studio’s built-in reports don’t show you. Feel free to copy-paste and use them in your Claude chat:

Video performance deep-dive

Prompt:

Analyze my last 20 videos. Group them by format (tutorial, review, opinion, Shorts). For each group, calculate the average view count, average watch time percentage, and average subscriber gain per video. Which format is delivering the best results?

Audience retention analysis

Prompt:

Look at my videos from the past 60 days. Which ones have the highest audience retention rate? What do they have in common — length, topic, or format? Identify any videos where retention drops significantly in the first 30 seconds and suggest possible reasons.

CTR vs. watch time gap analysis

Prompt:

Find videos where my click-through rate is above average but average view duration is below average. This suggests the title or thumbnail is working but the content isn't delivering on the promise. List these videos and suggest how I might improve content alignment.

Traffic source strategy

Prompt:

Break down my views by traffic source over the last 90 days — YouTube search, suggested videos, browse features, external, and direct. Which source is growing fastest? Which source delivers the highest average view duration? Recommend where I should focus my optimization efforts.

Subscriber growth attribution

Prompt:

Which specific videos drove the most subscriber gains in the last 30 days? Is there a pattern in the topics or formats that consistently convert viewers into subscribers vs. videos that get views but don't grow my channel?

Content calendar planning

Prompt:

Based on my last 6 months of YouTube data, what are my three best-performing content themes? Suggest a 4-week content calendar with specific video ideas that build on what's already working, with a mix of formats to test.

🤖 Browse the full Windsor AI Prompt Library for more ideas: https://windsor.ai/prompt-library/.

Advanced use cases: what you can do with YouTube Analytics in Claude

1. Automated weekly channel reports

Ask Claude to generate a structured weekly performance summary (top videos, subscriber growth, watch time trends, and one recommended action) and get a client-ready report in seconds without opening YouTube Studio.

2. Content strategy optimization

Feed Claude 3–6 months of video data and ask it to identify your channel’s strongest content pillars. Use those insights to build a data-driven content strategy rather than relying on gut feel or copying trends.

3. A/B title and thumbnail analysis

If you run title or thumbnail experiments, stream the performance delta into Claude and ask it to identify which variables (length, question format, numbers, emotional triggers) correlate with higher CTR across your catalog.

4. Cross-platform revenue attribution

Combine YouTube Analytics with Google Ads or GA4 data in the same Claude chat using Windsor-integrated data. Ask Claude whether your YouTube content drives measurable website traffic or ad-assisted conversions, connecting organic content performance to business outcomes.

5. Shorts vs. long-form strategy

With YouTube pushing Shorts heavily, use Claude to objectively compare your Shorts and long-form performance across subscriber gain, impressions, and watch time, and decide where your production effort is best invested.

YouTube Analytics data you can stream to Claude via Windsor

Windsor gives Claude access to a comprehensive set of YouTube Analytics fields, including the following:

Channel-level metrics

• Views — Total channel views over any date range
• Watch time (minutes) — Aggregate minutes watched, the signal YouTube’s algorithm weights most heavily
• Subscribers gained/lost — Net subscriber changes over time
• Estimated revenue — For monetized channels: AdSense revenue estimates
• Impressions — How often your thumbnails appeared on YouTube
• Impressions click-through rate (CTR) — The ratio of impressions that became views

Video-level performance

• Video ID & title — Unique identifier and title for each piece of content
• Average view duration — Mean seconds watched per view
• Average percentage viewed — What fraction of each video’s viewers actually watch
• Likes & dislikes — Engagement signals per video
• Comments — Comment count, reflecting audience interaction depth
• Shares — How often a video is shared off-platform
• Cards click rate — Engagement with end-screen and card CTAs

Audience & traffic sources

• Traffic source type — YouTube search, suggested, browse, external, notifications, and more
• Traffic source detail — Specific search queries driving traffic to your videos
• Viewer geography — Country and region breakdown of your audience
• Viewer demographics — Age group and gender of viewers (where available)
• Device type — Mobile, desktop, tablet, TV — critical for format and length decisions

Monetization metrics (for eligible channels)

• Estimated ad revenue — Revenue attributed to ads
• CPM (cost per mille) — Revenue per 1,000 ad impressions
• RPM (revenue per mille) — Total revenue per 1,000 views, including all sources
• Monetized playbacks — Views that included at least one ad

💡 Windsor normalizes all of these fields into a consistent, LLM-optimized schema so Claude can compare metrics across date ranges and video types without misinterpreting the data.

The manual way to get YouTube data into Claude (and why it fails)

Most creators and marketers start with the manual approaches to integrate YouTube data into Claude but face the same ceiling.

Manual method 1: YouTube Studio CSV export

YouTube Studio lets you export basic analytics as CSV files.

The typical workflow: download the report, open it in a spreadsheet, clean the data, then upload it to Claude. It works once. But:

• Every reporting cycle requires repeating the same export–clean–upload process.
• YouTube exports are limited to preset report types — you can’t get a single file combining video performance, traffic sources, and demographics.
• The exported data loses context: video titles are sometimes truncated, and there’s no way to include thumbnail URLs or descriptions for qualitative analysis.
• Data is stale the moment you export it; there’s no live feed.

Manual method 2: YouTube Studio screenshots or copy-paste

For quick checks, some users screenshot Studio dashboards or copy-paste stats into Claude. This is fine for one-off questions, but completely impractical for any regular reporting or multi-video analysis.

Manual method 3: YouTube Analytics API (for developers)

Developers can query the YouTube Analytics API directly. But this requires setting up OAuth, managing API quotas, handling pagination, and normalizing the response format before Claude can work with it reliably.

It’s a significant engineering investment that still needs ongoing maintenance as the API evolves.

Comparison: Top methods to connect YouTube Analytics to Claude

Conclusion

YouTube Analytics is packed with signals that most creators and marketers never fully act on, not because the data isn’t there, but because extracting and interpreting it manually is too slow to be useful. By the time you’ve exported, cleaned, and analyzed a report, the moment to act has often passed.

Windsor MCP changes that equation entirely. A one-time, 60-second setup gives Claude a permanent, live window into your channel, so you can ask any question, at any time, and get an answer based on your most current data.

Whether you’re a solo creator optimizing content strategy, a brand manager running a YouTube presence, or an agency reporting to clients, the ability to query your channel in plain English, without touching a spreadsheet, is a genuine competitive advantage.

🚀 Ready to get started? Begin your 30-day free trial at Windsor.ai and connect YouTube Analytics to Claude right now: https://onboard.windsor.ai/app/youtube.

FAQs

What are the main ways to connect YouTube Analytics to Claude?

There are three main approaches, each with different trade-offs:

• Windsor MCP (recommended): A no-code, native connector that creates a live, permanent pipeline between YouTube Analytics and Claude. Set up automatically in under 60 seconds, no maintenance required, always-fresh data.
• Manual CSV export: Download a report from YouTube Studio and upload it to Claude as a file. Works for one-off analysis but requires repeating the process every time, and only provides a static snapshot.
• Direct YouTube Analytics API: Build a custom integration using Google’s API. Gives full control but requires developer resources, OAuth setup, quota management, and ongoing maintenance.

For most creators, marketers, and agencies, Windsor MCP is the only approach that’s both fast enough to use regularly and deep enough to power real strategic analysis.

What is the fastest way to get YouTube Analytics into Claude?

Windsor MCP is by far the fastest method — the entire setup takes under 60 seconds, and you only do it once. After that, your YouTube data is permanently available in every Claude conversation with no additional steps. Compare that to a manual CSV export (10–20 minutes per reporting cycle) or a direct API build (days of engineering work).

Does the Windsor MCP-powered setup work with both personal YouTube channels and brand accounts?

Yes. Windsor connects to any YouTube channel you have analytics access to: personal creator accounts, brand accounts, and multi-channel networks. For agencies managing multiple client channels, you can connect several accounts within Windsor and query them separately or compare them within Claude.

What YouTube Analytics reports does Windsor support?

Windsor provides access to channel-level metrics, video-level performance, traffic source breakdowns, audience demographics, device data, and monetization metrics for eligible channels. See the full field list at windsor.ai/data-field/youtube/.

Can Claude access individual video performance, or only channel totals?

Both. Windsor streams data at both the channel level (aggregate views, watch time, subscriber trends) and the video level (per-video performance across all available metrics). This lets Claude compare individual videos, identify top performers, and spot outliers in a single query.

How far back can I query historical YouTube data?

Windsor can pull data from YouTube Analytics going back as far as your channel history allows. YouTube typically stores analytics data from a channel’s creation date. You can ask Claude to analyze trends across any custom date range.

Can I see which YouTube search terms are driving views?

Yes. Windsor includes traffic source detail data, which shows the specific search queries that led viewers to your videos. Ask Claude to identify your top-performing search terms and whether any high-volume queries are underserved by your current content.

Does Windsor support YouTube Shorts analytics separately?

Yes. YouTube Shorts data is included in the Windsor connector and can be filtered separately by video type. You can ask Claude to compare Shorts vs. long-form performance across any metric (views, subscriber gain, impressions, CTR) to build a data-driven format strategy.

Can I combine YouTube Analytics with other data sources in Claude?

Yes, this is one of Windsor’s most powerful features. You can blend YouTube Analytics with Google Ads, GA4, Shopify, or any of Windsor’s 325+ connectors in a single Claude conversation. For example: ‘Did the spike in YouTube views last Tuesday correlate with an increase in website conversions?’

Is any coding required to set this up?

No. The standard setup uses Windsor’s no-code interface and Claude’s built-in connector system. The developer configuration option (using mcp-proxy and a JSON config file) is available for technical users who prefer it, but it’s entirely optional.

How is this different from just uploading a YouTube CSV to Claude?

A CSV is a frozen snapshot — it’s outdated before you’ve finished the analysis. Windsor MCP is a live connection: every time you ask Claude a question, it fetches the latest data from YouTube’s API. You also get richer data in Windsor, including fields that YouTube’s standard exports don’t include in a single download.

Is Windsor MCP free to use?

Windsor MCP is included on all Windsor.ai plans, including the free 30-day trial and the Forever Free plan. See full details at windsor.ai/pricing/.