The “AI SDR” Agent (Inbound)

Build an autonomous AI agent that instantly responds to inbound leads, qualifies them through natural conversation, and books meetings for sales reps.

Advanced Complexity
Owner: RevOps / Engineering
Updated Jan 2025
Workflow overview diagram

Workflow overview

Download diagram

Trigger

New inbound lead (form fill, demo request, chat)

Inputs

Lead form data, enrichment data, qualification criteria, rep availability

Output

Qualified lead handoff, booked meeting, nurture sequence assignment

Success Metrics

Speed to lead <2 min, qualification accuracy >90%, meeting book rate >40%

Overview

What It Is

An AI-powered system that responds to inbound leads within minutes, engages in natural qualification conversation via email or chat, determines fit against ICP criteria, and either books meetings with sales or routes to appropriate nurture.

Why It Matters

Speed to lead is critical—responding within 5 minutes vs. 30 minutes increases qualification rates 21x. AI SDR enables instant response 24/7, ensuring no hot lead goes cold while waiting for human follow-up.

Who It's For

  • Companies with high inbound volume
  • Teams with limited SDR capacity
  • Organizations selling globally across time zones
  • Growth teams optimizing lead conversion

Preconditions

Required Tools

  • GPT-4/Claude API (conversation)
  • HubSpot/Salesforce (CRM)
  • Calendly/Chili Piper (booking)
  • Clearbit/ZoomInfo (enrichment)
  • n8n (orchestration)

Required Fields/Properties

  • Lead form fields (name, email, company)
  • Qualification criteria (BANT or similar)
  • Sales rep availability calendars
  • Routing rules by segment/territory

Definitions Required

  • Qualified lead criteria
  • Disqualification criteria
  • Routing rules (which leads to which reps)
  • Escalation triggers to human SDR

Step-by-Step Workflow

1

Configure Instant Enrichment

Goal: Gather context before first outreach

Actions:

  • Trigger enrichment on form submission
  • Pull company data (size, industry, tech stack)
  • Get contact data (title, LinkedIn, phone)
  • Pre-score lead based on ICP fit

Implementation Notes: Enrichment must complete in <30 seconds to enable fast response. Use waterfall enrichment (Clearbit → ZoomInfo → Apollo) to maximize coverage.

2

Build Qualification Conversation Agent

Goal: Create AI that qualifies leads through natural dialogue

Actions:

  • Define qualification questions (BANT, MEDDIC, etc.)
  • Build conversational qualification prompt
  • Handle multi-turn conversations
  • Extract structured qualification data

Implementation Notes: Qualification should feel like helpful conversation, not interrogation. Lead with value, weave in questions naturally, adapt based on responses.

3

Implement Multi-Channel Response

Goal: Enable AI to engage via email, chat, or SMS as appropriate

Actions:

  • Email response for demo requests
  • Chat widget integration for website visitors
  • SMS option for mobile leads
  • Consistent AI personality across channels

Implementation Notes: Match channel to lead preference. If they filled a form at 10pm, email. If they're on site now (chat widget), engage in real-time.

4

Create Smart Booking Flow

Goal: Enable AI to book meetings based on routing rules

Actions:

  • Define routing rules (territory, segment, round-robin)
  • Integrate with rep calendars
  • Handle timezone detection
  • Confirm and send calendar invites

Implementation Notes: Route based on territory first, then segment specialization. If assigned rep has no availability in 3 days, fall back to round-robin.

5

Build Handoff to Human SDR

Goal: Seamless transition when human needed

Actions:

  • Define human escalation triggers
  • Create handoff summary for SDR
  • Maintain conversation context
  • Notify SDR with full context

Implementation Notes: Human escalation for: enterprise leads, complex technical questions, competitor mentions, expressed frustration, or AI uncertainty >30%.

6

Configure Nurture Routing

Goal: Appropriately handle leads that don't qualify for sales

Actions:

  • Define nurture criteria and segments
  • Build nurture sequence assignment
  • Enable re-qualification triggers
  • Track nurture progression

Implementation Notes: Not all inbounds are sales-ready. Route to appropriate nurture based on disqualification reason: timing → drip campaign, company size → SMB self-serve, etc.

Templates

Initial Response Email

Subject: Thanks for reaching out, {{first_name}}!

Hi {{first_name}},

Thanks for your interest in {{product}} - I saw you're looking to {{inferred_goal}}.

Given that {{company}} is {{company_context}}, I think you'd find our {{relevant_feature}} particularly valuable. {{similar_company}} saw {{result}} after implementing it.

A few quick questions to make sure I point you in the right direction:
- What's prompting you to explore this now?
- What tools are you currently using?

Happy to share some relevant resources or jump on a quick call - whatever's most helpful.

Best,
{{rep_name}}

Handoff Summary Template

**AI SDR Handoff Summary**

**Lead:** {{name}} - {{title}} at {{company}}
**Lead Score:** {{score}}/100
**Escalation Reason:** {{reason}}

**Conversation Summary:**
{{conversation_summary}}

**Qualification Status:**
- Pain: {{pain_identified}}
- Authority: {{authority_level}}
- Need: {{need_urgency}}
- Timeline: {{timeline}}

**Key Insights:**
{{key_insights}}

**Suggested Next Step:**
{{recommended_action}}

**Full Conversation History:**
[Link to conversation thread]

Inbound AI SDR Metrics

| Metric | Target | This Week | Status |
|--------|--------|-----------|--------|
| Avg Response Time | <5 min | 2.3 min | On Track |
| Response Rate | 100% | 100% | On Track |
| Qualification Rate | >30% | 34% | On Track |
| Meeting Book Rate | >40% | 42% | On Track |
| Human Escalation | <15% | 12% | On Track |
| Lead Satisfaction | >4.0 | 4.2 | On Track |

QA + Edge Cases

Test Cases Checklist

  • Verify response sent within 5 minutes of form submission
  • Test qualification conversation handles multi-turn dialogue
  • Confirm routing rules assign correct reps
  • Validate calendar booking creates correct events
  • Test human escalation triggers work correctly

Common Failure Modes

  • Delayed response: Enrichment timeout delays initial email. Use fallback data and respond immediately, enrich in background.
  • Over-qualification: AI asks too many questions, feels like interrogation. Limit to 2-3 key questions, spread across messages.
  • Wrong rep routing: Lead assigned to unavailable or wrong-fit rep. Add availability check and segment matching to routing.

Troubleshooting Tips

  • If meeting book rate is low, review qualification email tone
  • For high escalation rate, expand AI's objection handling
  • If leads complain about AI, add more natural language variation

KPIs and Reporting

KPIs to Track

  • Speed to Lead: <5 minutes first response
  • Qualification Rate: >30% of inbounds qualified
  • Meeting Book Rate: >40% of qualified leads book
  • Lead Satisfaction Score: >4.0 / 5.0

Suggested Dashboard Widgets

  • Response Time Distribution: Histogram of time to first response
  • Qualification Funnel: Inbound → Engaged → Qualified → Meeting → Opportunity
  • AI vs Human Performance: Compare conversion rates AI-handled vs human
  • Channel Mix: Inbound by source and AI performance per source

Want This Implemented End-to-End?

If you want this playbook configured in your stack without the learning curve:

  • Timeline: Week 1: Enrichment + initial response. Week 2: Qualification + booking. Week 3: Testing + launch.
  • Deliverables: Inbound AI SDR system, routing rules, monitoring dashboard
  • Handoff: Engineering builds, RevOps configures rules, SDRs handle escalations
Request Implementation
Jump to Steps Implement