Google Maps Lead Generation (MCP)

Use this skill to run repeatable lead-gen batches from Google Maps via MCP.

Preconditions

• mcporter configured with server google-maps.
• Server key in env (GOOGLE_MAPS_API_KEY) must be server-compatible (no browser referrer restriction).
• For XLSX output, openpyxl available in venv.

Fast workflow

1. Build query set from geography + target verticals.
2. Run maps_search_places for each query.
3. Keep only in-target geography, dedupe by place_id.
4. Enrich each place with maps_place_details.
5. Export CSV or XLSX.
6. If user asks for file in Telegram, send with message tool action=send + media path.

Query strategy

Use focused terms instead of broad generic terms.

• Good: "odoo partner <city> <country>", "erp integrator <city> <country>", "logistics company <city> <country>"
• Avoid huge overlapping lists in one run; do batches.

Required output columns (V2)

• name
• address
• phone
• website
• email (empty if not discoverable)
• rating
• place_id
• google_maps_url (mobile-safe):
  • https://www.google.com/maps/search/?api=1&query=<NAME>&query_place_id=<PLACE_ID>

Cost notes

• Search calls are usually main paid SKU driver.
• Place details add enrichment cost.
• Report rough run cost estimate and mention free-tier caveat.

Reliability guardrails

• Batch enrich in small chunks (10–50) to avoid long-running timeouts.
• Add retries for transient failures.
• Never commit API keys or sensitive exports.

Delivery rules

• If user asks for CSV/XLSX file in chat: send via message tool (media path).
• If user asks specifically for XLSX formatting/edits, use xlsx workflow standards.
• Keep summary concise: count, coverage (with_phone, with_website, with_email), file path/name.