The OpenClaw analytics environment has been significantly upgraded. Beyond GA4, Microsoft Clarity and Google Search Console (GSC) MCP servers have been added, enabling multi-dimensional analysis from three data sources.

Two New MCP Servers

1. Microsoft Clarity MCP (@microsoft/clarity-mcp-server)

Clarity is a free heatmap and session recording tool. Via the MCP server, the following data is accessible:

  • Dead Click / Rage Click: Locations where clicks produce no response, or where users click repeatedly in frustration
  • Scroll Depth: How far down the page users read
  • Device-Specific Engagement: Behavioral differences between desktop and mobile
  • Session Recordings: Actual user behavior captured as video

The Clarity tag (Tag ID 41) was already configured in GTM, so adding the MCP server immediately enabled analysis.

2. Google Search Console MCP (mcp-server-gsc)

GSC provides search performance data. Up to 25,000 rows can be retrieved:

  • Top Search Queries: Keywords driving discovery
  • Page-Level Performance: Impressions, clicks, CTR, position
  • CTR Improvement Candidates: High-impression, low-click pages
  • Index Status: Sitemap and crawl information

Setup Procedure

Adding via mcporter

mcporter add clarity
mcporter add gsc

Two MCP servers added with these commands.

Authentication Configuration

Clarity Authentication

Clarity uses per-project authentication. Complete the auth flow in a browser.

GSC Authentication

The same service account used for GA4 can be reused:

  1. Enable Search Console API in GCP
  2. Add the service account in Search Console
    • Search Console → Settings → Users and permissions
    • Add the service account email with “Owner” permission
  3. Environment variable (skip if already set for GA4) 
    export GOOGLE_APPLICATION_CREDENTIALS=/path/to/service-account.json

Authentication complete. GA4, GSC, and Clarity all accessible from the same OpenClaw instance.

Quick Wins from GSC Integration

Reviewing GSC data immediately revealed actionable improvements:

1. “banananl” Query Optimization

  • Impressions: 193
  • CTR: Only 5%
  • Position: 6.5

Appearing in search results but not getting clicked. Adding “BananaNL” explicitly to the meta title could significantly improve CTR.

2. “banana x prompt pattern collection” Refinement

  • Impressions: 24
  • CTR: 21% (high)
  • Position: 2.8

Already ranking well with strong CTR. Further title tag optimization could push to position 1.

Expanding the Cron Job to Three-Pillar Integrated Analysis

The existing “bananaX Daily GA4 Report” cron job was updated to integrate all three data sources:

Changes

  • Data sources: GA4 only → GA4 + Clarity + GSC
  • Timeout extension: 300s → 420s (sequential API calls to three services)
  • Execution time: 6:00 AM JST daily (21:00 UTC)

Starting the next morning, automated analysis runs in this sequence:

  1. GA4: Page views, events, conversions
  2. Clarity: Heatmaps, dead clicks, scroll depth
  3. GSC: Search queries, CTR, position

Cross-referencing three data sources enables early detection of issues like “appearing in search but not getting clicked” or “getting clicked but immediately bouncing.”

Summary

Adding Clarity MCP and GSC MCP to OpenClaw advances analytics capabilities to the next level:

  • GA4: Understand “what happened” through metrics
  • Clarity: Understand “why it happened” through user behavior observation
  • GSC: Understand “how users found the site” through search data

With three data sources in place, the foundation for a data-driven improvement cycle is complete.