5 Search Console Queries You Can Ask Claude with MCP (Copy-Paste Prompts)

Forget clicking through Google Search Console dashboards. With MCP (Model Context Protocol), you can query your GSC data using natural language—and get actionable insights in seconds.

This guide shows you exactly what to ask Claude when you connect a GSC MCP server. Each query includes the exact prompt, what Claude returns, and why it matters for your SEO strategy.

Why Natural Language GSC Queries Change Everything

Traditional GSC analysis requires:

• Navigating multiple dashboard tabs

• Setting up filters manually

• Exporting to spreadsheets for cross-referencing

• Repeating the process for different date ranges

With a GSC MCP server, you just ask Claude what you want to know. The MCP server handles the API calls, data aggregation, and formatting—you get the insights.

Let's look at five high-impact queries that SEOs use daily.

---

Query 1: Find Pages with High Impressions but Low CTR

The Prompt:

Show me pages with high impressions but low CTR over the last 28 days. 
Sort by impressions descending and show click-through rate.

What Claude Returns:

Claude queries your GSC data and returns a table like:

Page	Impressions	Clicks	CTR	Avg Position
/blog/seo-guide	15,420	312	2.02%	8.3
/features/analytics	12,890	198	1.54%	11.2
/pricing	8,760	245	2.80%	6.1

Why It Matters:

This is the classic "quick wins" analysis. Pages with high impressions but low CTR have visibility—Google is showing them—but something's preventing clicks. Common fixes:

• Title tag optimization: Make it more compelling or add power words

• Meta description improvement: Add a clear benefit or call-to-action

• Rich snippet opportunities: Add FAQ schema, how-to markup, or review stars

A page ranking at position 8 with 15,000 impressions and 2% CTR could realistically hit 5-6% CTR with better SERP presentation. That's an extra 450-600 clicks per month without improving rankings at all.

---

Query 2: Find Keywords Driving Traffic to a Specific Page

The Prompt:

Which keywords bring traffic to my /blog/ pages? 
Show me the top 20 keywords by clicks.

What Claude Returns:

{
  "page": "/blog/",
  "period": "Last 28 days",
  "keywords": [
    { "keyword": "seo tools 2025", "clicks": 892, "impressions": 12450, "ctr": "7.16%", "position": "4.2" },
    { "keyword": "best keyword research", "clicks": 654, "impressions": 8920, "ctr": "7.33%", "position": "5.1" },
    { "keyword": "how to improve rankings", "clicks": 423, "impressions": 15680, "ctr": "2.70%", "position": "9.8" }
  ]
}

Why It Matters:

Understanding which keywords drive traffic to specific pages helps you:

1. Avoid keyword cannibalization: If the same keyword drives traffic to multiple pages, you might be competing with yourself

2. Identify content gaps: High-impression, low-CTR keywords on a page might deserve their own dedicated article

3. Optimize existing content: Add sections that better address top-performing queries

This query is especially powerful for blog content audits. You can quickly see which topics resonate and which pages might need consolidation or expansion.

---

Query 3: Compare This Month vs Last Month Performance

The Prompt:

Compare my search performance for the last 7 days versus the previous 7 days.
Show total clicks, impressions, and average CTR for each period.

Pro Tip: Use the days Parameter

One quirk with LLMs is they can default to outdated dates from their training data. A query like "show me data from November 2024" might confuse the model about the current date.

Ekamoira's GSC MCP server solves this with a days parameter that calculates dates server-side:

Get my search analytics for the last 7 days (use days=7, not specific dates)

This ensures you always get data relative to today, not a date the model hallucinated.

What Claude Returns:

Period	Clicks	Impressions	CTR	Avg Position
Dec 24-31	4,520	89,430	5.05%	12.3
Dec 17-24	4,180	85,670	4.88%	12.8
Change	+8.1%	+4.4%	+0.17pp	-0.5

Why It Matters:

Weekly and monthly comparisons help you:

• Spot algorithm updates: Sudden drops across many pages often indicate a Google update

• Measure content launches: Did that new blog post you published actually move the needle?

• Track seasonal patterns: Some industries see predictable weekly/monthly cycles

Without MCP, this comparison requires two separate GSC exports and manual calculation. With Claude, it's one natural language query.

---

Query 4: Identify Quick-Win Keywords (Positions 4-10)

The Prompt:

Find keywords where I rank between positions 4 and 10 with at least 100 impressions.
Sort by impressions descending. These are my quick-win optimization opportunities.

What Claude Returns:

{
  "period": "Last 28 days",
  "quickWins": [
    { "keyword": "mcp server tutorial", "position": "6.2", "impressions": 3420, "clicks": 156, "ctr": "4.56%" },
    { "keyword": "claude api examples", "position": "7.8", "impressions": 2890, "clicks": 89, "ctr": "3.08%" },
    { "keyword": "seo automation tools", "position": "5.4", "impressions": 2450, "clicks": 178, "ctr": "7.27%" }
  ]
}

Why It Matters:

Keywords ranking 4-10 are your highest-ROI optimization targets:

• Already page one: Google considers your content relevant

• Just below the fold: Moving from position 7 to position 3 can 3-4x your clicks

• Easier than new content: Optimizing existing content is faster than creating new pages

For each quick-win keyword, consider:

• Adding internal links from higher-authority pages

• Expanding the content section that targets this keyword

• Improving page speed and Core Web Vitals

• Adding schema markup for rich snippet opportunities

---

Query 5: Check URL Indexing Status

The Prompt:

Is this URL indexed: example.com/my-new-page

Check the indexing status and tell me if there are any issues.

What Claude Returns:

{
  "url": "https://example.com/my-new-page",
  "indexStatus": {
    "verdict": "PASS",
    "coverageState": "Submitted and indexed",
    "robotsTxtState": "ALLOWED",
    "indexingState": "INDEXING_ALLOWED",
    "lastCrawlTime": "2025-12-28T14:23:00Z",
    "pageFetchState": "SUCCESSFUL",
    "crawledAs": "MOBILE"
  },
  "mobileUsability": "PASS"
}

Why It Matters:

Before you spend time optimizing a page, you need to know if Google can even see it. The URL Inspection API reveals:

• Is the page indexed? No point optimizing unindexed content

• Are there crawl blocks? Robots.txt or noindex issues

• Mobile usability: Critical for mobile-first indexing

• Last crawl time: How fresh is Google's version?

This query is essential after:

• Publishing new content

• Fixing technical issues

• Redirecting old URLs

• Recovering from a manual action

---

Getting Started with GSC MCP

Ready to try these queries yourself? You have two options:

Option 1: Hosted GSC MCP (Zero Setup)

Ekamoira's GSC MCP server works with Claude.ai, ChatGPT, and Cursor out of the box:

1. Connect your Google Search Console account

2. Add the MCP endpoint to your AI assistant

3. Start querying with natural language

The hosted version includes:

• Automatic token refresh

• Server-side date calculation (no LLM date confusion)

• Fresh data with dataState: 'all' (matches GSC dashboard)

• 30-day free trial

Option 2: Self-Hosted GSC MCP

If you prefer running your own server, check out:

• AminForou/mcp-gsc (Python)

• ahonn/mcp-server-gsc (TypeScript)

Self-hosted requires:

• Google Cloud OAuth setup

• Local credential management

• Node.js or Python environment

---

More GSC Prompts to Try

Once you're connected, experiment with these additional queries:

• "List all my sitemaps and their status"

• "Show me my top 10 keywords by impressions"

• "Which pages rank for [competitor keyword]?"

• "Test if example.com/page is mobile-friendly"

• "Submit my new sitemap at example.com/sitemap.xml"

The power of MCP is that you're not limited to predefined reports. Ask whatever question you have about your search performance, and Claude will figure out which API calls to make.

---

Key Takeaways

1. High impressions + low CTR = Quick-win title/meta optimization opportunities

2. Page-level keyword analysis = Understand what's actually ranking (and why)

3. Period comparisons = Track progress and spot algorithm updates

4. Position 4-10 keywords = Your highest-ROI optimization targets

5. URL inspection = Verify indexing before wasting optimization effort

Stop clicking through dashboards. Start asking Claude questions about your search data.

Get started with GSC MCP →