Used byA big security guard training provider uses this internally to find low-rated, training-deficient local security companies as warm outbound targets.
Built withDeveloped using growth-mcp, the open MCP server that powers our GTM workflows.
1
Seeded ~5k US places (cities, towns, neighborhoods) from US Census gazetteer + population ranks.
2
Searched Google Maps for "security guard" + nearby variants in every place via Apify, deduped by place_id.
3
Classified each business. Is this actually a security guard service, or a hardware store / locksmith / alarm installer?
4
Fetched reviews for every classified guard company, focusing on ≤2★ negatives.
5
Labeled training issues. A Claude sub-agent reads each negative review and flags untrained, unprofessional, or absent-on-shift guards with evidence.
6
Tagged brand & franchise. Normalized brand names and detected national chains (≥5 locations under the same canonical brand).