Summary
Nate Herk walks through a complete lead generation workflow combining Claude Code as an orchestration layer with Clay as a B2B data source, demonstrating how to produce 50 enriched leads with personalized cold email copy using only natural language prompts — no manual tool-switching or UI navigation required.
The workflow begins by connecting Clay to Claude Code via API, then issuing a goal prompt (/goal flag) that specifies the desired end state: in this case, 50 HVAC and home-service business owners with verified email addresses, company pain points, recent notable achievements, and individually personalized subject lines and email bodies. Claude Code autonomously navigates Clay’s enrichment waterfall — which queries multiple B2B data providers in sequence to maximize email-find rates — and compiles everything into a CSV. Context files loaded into the Claude Code workspace (business profile, case studies, offer details, FAQs, and website copy) allow the model to write outreach grounded in specific service positioning rather than generic AI-sounding copy.
Herk covers the full setup process including Clay account connection, the structure of Clay’s MCP action library (find emails, enrich person/company, look up phone), the use of /goal to keep Claude Code running until all 50 leads meet spec, and real credit costs from his demo run. The tutorial targets solo operators and small agencies looking to run cold outbound at scale without managing separate subscriptions for data enrichment, copywriting, and campaign tooling.
📺 Source: Nate Herk | AI Automation · Published July 12, 2026
🏷️ Format: Tutorial Demo
