LeadRescuePro Full AI SOP

Effective Date: February 26, 2026

This SOP defines the production operating standard for AI-driven lead intake, qualification, booking, escalation, and quality control.

1. Scope

  • Inbound missed-call capture and AI handoff.
  • ZIP-first qualification and practical intake.
  • Booking/callback outcomes and statusing.
  • Lead log and dashboard write-through.
  • Daily operator review and QA.
  • Billing gate and cancellation detach policy.

2. Non-Negotiable Rules

  • Ring-first: customer phone rings first for 10 seconds.
  • AI handles only missed/unanswered calls.
  • ZIP is collected before any full intake.
  • Out-of-area leads are politely declined and logged.
  • Address is captured only after qualification and booking intent.

3. AI Workflow

  1. Inbound call arrives and rings customer for 10 seconds.
  2. If answered by human, AI path is skipped.
  3. If missed/no-answer/busy, AI takes over.
  4. AI validates ZIP and service area.
  5. In-area calls continue to intake and booking.
  6. Outcome is stored as booked, not booked, or follow-up needed.

4. Required Lead Fields

  • Lead ID, timestamp, caller phone, caller name (if provided).
  • ZIP code and in-area qualification result.
  • Issue summary and urgency label.
  • Outcome status and appointment timestamp (when booked).
  • Transcript excerpt and source reference for idempotency.

5. Status Contract

Statuses must be one of: missed, answered_by_plumber, answered_by_ai, booked, not_booked, followup_needed, closed.

6. Escalation Rules

  • Emergency and safety-risk calls are prioritized and escalated immediately.
  • Caller request for human triggers warm transfer if available, otherwise urgent callback task.
  • High-urgency leads are surfaced to the top of operator queue.

7. Daily Operator SOP

  1. Review new leads in the morning and late afternoon.
  2. Process in this order: emergency, high, booked confirmations, follow-up needed.
  3. Update pipeline state after each customer touchpoint.
  4. Close stale leads with reason code.

8. QA and Reliability

  • Daily QA sample: 10 random AI-handled calls.
  • Failure criteria: wrong qualification path, missing required fields, incorrect urgency, hallucinated details.
  • If daily QA failure exceeds 5%, move to safe prompt profile and perform prompt review.

9. Billing and Service Gate

  • AI workflow only runs for eligible setup and subscription states.
  • Usage billing uses idempotent references to prevent duplicate charging.
  • Service gate blocks AI actions when account is not eligible.

10. Cancellation and Auto-Detach

  • Cancellation timestamp is recorded when cancellation is requested.
  • After 24 hours, service is auto-detached.
  • Detach action disables service and clears AI bindings.
  • After detach, AI fallback path hard-stops on call handoff.