Ghost Engine

Appearance

Geogrid Scan Workflow ​

Fresh

Source: ROADMAP.md (6.1 Geogrid scanner), BIBLE.md (Phase 12 Phase 6)

Map pack rank scanning across lat/lng grid points (neighborhoods or zip codes). Heatmap visualization on dashboard.

Steps ​

1. Pick a center and radius ​

The grid is anchored on a lat/lng with a radius in miles.

Example: Dallas at (32.7767, -96.7970) with 15-mile radius.

2. Pick a grid size ​

A 7x7 grid covering 15 miles produces 49 sample points spaced ~3.75 miles apart.

A 9x9 grid for the same area: 81 points, ~3.3 miles apart.

3. Run the scan ​

bash
curl -X POST http://localhost:4700/geogrid/scan \
  -H "Content-Type: application/json" \
  -H "x-api-key: ghost-engine-key" \
  -d '{
    "keyword": "junk removal",
    "domain": "journeyjunkremoval.com",
    "center": { "lat": 32.7767, "lng": -96.7970 },
    "radiusMiles": 15,
    "gridSize": 7
  }'

At each grid point, the scanner:

  • Sets the browser geo to that lat/lng
  • Searches Google Maps for the keyword
  • Records the position of the target domain in the map pack
  • Records null if not in top 20

4. View the heatmap ​

bash
curl http://localhost:4700/geogrid/scans/geo-abc123

Response includes a 2D matrix of ranks. The dashboard renders this as a heatmap with color-coded cells.

5. Compare to a previous scan ​

bash
curl "http://localhost:4700/geogrid/compare?from=geo-abc123&to=geo-def456"

Returns per-point delta and a "moved up/down/same" summary.

6. Schedule recurring scans ​

bash
curl -X POST http://localhost:4700/geogrid/schedule \
  -H "Content-Type: application/json" \
  -d '{
    "keyword": "junk removal",
    "domain": "journeyjunkremoval.com",
    "center": { "lat": 32.7767, "lng": -96.7970 },
    "radiusMiles": 15,
    "gridSize": 7,
    "frequency": "weekly"
  }'

What the scan reveals ​

  • Where you are strong (high-rank quadrants)
  • Where you are weak (unranked quadrants)
  • Geographic edges of your authority
  • Drift over time (compare scans)

Cost considerations ​

A 7x7 = 49 SERP scrapes per scan. A 9x9 = 81 scrapes. Weekly cadence is normal; daily on high-value keywords only.