How to Track Monthly SEO Delta Changes for a Domain Using geo-seo-claude

Use the geo-compare skill to compare two GEO-SEO audit reports and automatically compute score differences, trend symbols, and actionable monthly insights.

Tracking monthly SEO delta changes for a domain requires comparing baseline performance against current metrics to identify trends and measure optimization progress. The geo-seo-claude repository provides a dedicated workflow through the geo-compare skill, which automates the comparison of audit reports stored in ~/.geo-prospects/audits/ and generates client-ready progress documentation. This system enables data-driven SEO strategies by calculating precise deltas for GEO scores, AI citability, and platform readiness metrics.

Understanding the Monthly Delta Tracking Architecture

The delta tracking system operates through a structured file-based workflow defined in skills/geo-compare/SKILL.md. This architecture ensures that every audit is preserved as a historical record while enabling automated comparison between any two points in time.

Data Storage and File Structure

Audit reports are persisted as Markdown files in the ~/.geo-prospects/audits/ directory, using a naming convention that includes the domain and date (e.g., electron-srl.com-2026-01-15.md). When executing a comparison, the skill scans this directory and automatically selects the oldest file as the baseline and the newest as the current audit, or accepts explicit file paths as arguments. The resulting monthly delta report is written to ~/.geo-prospects/reports/ with a filename indicating the domain and comparison period.

Parsing and Scoring Mechanism

The skill parses structured data from audit files to extract critical performance indicators. Specifically, it locates lines containing GEO Score, AI Citability, and category-specific scores formatted as Metric: XX/100. Additionally, the parser gathers AI-crawler status, critical issues, and action-item progress to provide comprehensive context for the delta calculation.

How the Delta Calculation Works

The core mathematics behind tracking monthly SEO delta changes involves straightforward subtraction with intelligent trend analysis. For every extracted metric, the system calculates Delta = Current Value - Baseline Value.

The system assigns visual trend indicators to each delta:

  • ▲ (improved) when the current score exceeds the baseline
  • ▼ (declined) when the current score falls below the baseline
  • ── (unchanged) when no difference is detected

These symbols, combined with color-coded values, appear directly in the generated Markdown reports to provide immediate visual feedback on performance trajectories.

Generating Monthly Progress Reports

The geo-compare skill uses a predefined Markdown template to construct professional client deliverables. Each report contains:

  • Executive summary highlighting high-level delta changes and overall trajectory
  • Score breakdown tables displaying before/after values alongside calculated differences
  • Platform readiness changes showing which search platforms improved or degraded
  • AI-crawler access updates indicating changes in bot accessibility
  • Action-plan status tracking completion of previously recommended optimizations
  • 6-month projection estimating future performance based on current velocity

Running the Comparison Workflow

Execute the delta tracking workflow through the Claude Code CLI using the /geo compare command. The skill supports both automatic file detection and manual specification of audit files.

To automatically compare the oldest and newest audits for a domain:

/geo compare electron-srl.com

To compare specific audit files:

/geo compare ~/.geo-prospects/audits/electron-srl.com-2026-01-15.md ~/.geo-prospects/audits/electron-srl.com-2026-03-12.md

According to the source code in skills/geo-compare/SKILL.md, these commands trigger the full delta computation pipeline and output confirmation messages upon successful report generation.

For programmatic integration, invoke the comparison using Python's subprocess module:

import subprocess
from pathlib import Path

def run_geo_compare(domain: str):
    """Run the /geo compare command for a given domain."""
    cmd = ["claude", "code", "run", f"/geo compare {domain}"]
    result = subprocess.run(cmd, capture_output=True, text=True)
    if result.returncode != 0:
        raise RuntimeError(f"Geo compare failed: {result.stderr}")
    print(result.stdout)

# Example usage

run_geo_compare("electron-srl.com")

To target specific audit files programmatically:

subprocess.run([
    "claude", "code", "run",
    "/geo compare",
    str(Path.home() / ".geo-prospects/audits/electron-srl.com-2026-01-15.md"),
    str(Path.home() / ".geo-prospects/audits/electron-srl.com-2026-03-12.md")
])

These commands generate reports at paths like ~/.geo-prospects/reports/electron-srl.com-monthly-2026-03.md, confirming successful execution as documented in the skill implementation.

Integration with the Audit Pipeline

The monthly delta tracking system functions as the analytical component of a complete GEO-SEO workflow. The geo-audit skill (invoked via /geo quick or full audit commands) generates the raw Markdown files that geo-compare consumes. This creates a seamless operational loop: execute an audit to establish a baseline, store the file automatically in ~/.geo-prospects/audits/, run subsequent audits to capture current state, and invoke /geo compare to generate delta documentation. The architecture diagram in README.md (lines 12-31) illustrates this relationship, showing how audit generation feeds directly into comparison and reporting modules.

Summary

  • Store audits consistently: The system saves all audits to ~/.geo-prospects/audits/ as timestamped Markdown files, enabling historical comparison.
  • Use /geo compare: Invoke this command with either a domain name (auto-selects oldest/newest) or explicit file paths to trigger delta calculation.
  • Understand the math: Delta equals Current minus Baseline, with trend symbols (▲, ▼, ──) indicating direction of change.
  • Generate client reports: The skill produces comprehensive Markdown reports in ~/.geo-prospects/reports/ containing executive summaries, score tables, and projections.
  • Integrate with geo-audit: Run regular audits using the geo-audit skill to feed fresh data into the comparison pipeline.

Frequently Asked Questions

What file format does the monthly delta tracking system use?

The system uses standard Markdown files for both input and output. Audit files stored in ~/.geo-prospects/audits/ contain structured text with scores formatted as Metric: XX/100, while generated reports in ~/.geo-prospects/reports/ use Markdown tables and formatting for professional presentation.

Can I compare audits from specific dates rather than oldest vs. newest?

Yes. While the default behavior selects the oldest file as baseline and newest as current, you can specify exact file paths in the command: /geo compare path/to/baseline.md path/to/current.md. This allows precise comparison between any two historical audit points.

How does the system determine if a metric has improved or declined?

The system calculates the mathematical difference between current and baseline scores. Positive results display the ▲ symbol (improved), negative results show ▼ (declined), and zero differences render as ── (unchanged). This visual encoding appears alongside color-coded numerical values in the generated reports.

Where can I find the generated monthly progress report?

After successful execution, the skill writes the delta report to ~/.geo-prospects/reports/ using a filename pattern like [domain]-monthly-[date].md. The CLI output confirms the exact file path and location upon completion.

Have a question about this repo?

These articles cover the highlights, but your codebase questions are specific. Give your agent direct access to the source. Share this with your agent to get started:

Share the following with your agent to get started:
curl -s "https://instagit.com/install.md"

Works with
Claude Codex Cursor VS Code OpenClaw Any MCP Client

Maintain an open-source project? Get it listed too →