Outbound Sales Flow

This is the most common agent pattern. Your agent takes a lead, researches them, finds a meeting opportunity, dispatches a sales rep, and captures the outcome.

The pipeline

Email → Lead → Interests → Events/Places → Rep → Book → Invite → Report → Payment

Each step maps to one or more HumanConnection tools:

StepToolInput
Research the leadsearch_humansname, company_domain
Discover interestssearch_interestsname, linkedin_url, x_url
Find eventssearch_eventslocation, category
Find venuessearch_placeslocation, type
Find ticketssearch_ticketsevent_url
Find a sales repsearch_humanslocation, title_filters
Book the eventbook_ticketevent_url, email, name
Send invitesend_meeting_inviteattendee_emails, scheduled_at, location
Capture outcomesend_meeting_reportmeeting_id, outcome, deal_value
Pay the repsend_meeting_paymentmeeting_id, amount, type

Example: Spanx lead

Your agent receives a lead: Sara Blakely, Founder of Spanx.

  1. search_humans({ name: "Sara Blakely", company_domain: "spanx.com", provide_email: true }) — returns her LinkedIn, X, Instagram, location (Atlanta), and email
  2. search_interests({ name: "Sara Blakely", linkedin_url: "...", x_url: "..." }) — finds signals: philanthropy, fitness, entrepreneurship
  3. search_events({ location: { city: "Atlanta" }, category: "conference" }) — finds a retail leadership summit
  4. search_humans({ location: "Atlanta", title_filters: "sales AND retail" }) — finds a nearby rep
  5. book_ticket(...) + send_meeting_invite(...) — sets up the meeting
  6. send_meeting_report(...) + send_meeting_payment(...) — closes the loop

Tested scenarios

This pattern has been validated end-to-end across 5 industries:

LeadCompanyLocationKey interest signal
Sara BlakelySpanxAtlantaPhilanthropy, fitness
David SolomonGoldman SachsNew YorkDJ, electronic music
Ryan GellertPatagoniaVenturaSustainability, climbing
Lidiane JonesBumbleAustinWomen in tech
Melanie PerkinsCanvaSydneyDesign, education

Each scenario follows the same pipeline. The agent adapts the event category and venue type based on the interest signals discovered in step 2.

Decision points for your agent

  • Confidence < 0.4: The lead match is weak. Consider verifying manually before proceeding.
  • No events found: Fall back to search_places for a restaurant or cafe meeting.
  • No rep nearby: Expand the location radius or use a different title_filters query.
  • Lead has no email: Use provide_email: true in the initial search, or ask your CRM for the email.

Next steps