AI Lead Qualification at Scale
Replace manual SDR qualification calls with AI. This guide sets up a system that calls leads, asks qualifying questions, scores engagement, creates deals for hot prospects, and nurtures undecided contacts automatically.
The System You'll Build
End-to-end flow
New lead enters CRM (form fill, ad click, import)
↓ Auto-added to qualification campaign
↓ AI calls within configured schedule
↓ Asks about budget, timeline, decision authority
↓ Scores engagement 0–100
↓ Score 70+: Creates deal, advances to SQL, alerts sales team
↓ Score 40–70: Retries in 4–8 hours with adapted approach
↓ Score <40: Marks not interested, stops calling
↓ Campaign completes → AI report with insights
Step 1: Create the Qualification Agent
Go to Agents → New Agent:
System Prompt
You are a business development representative for [Your Company]. You're calling leads who expressed interest in our product. Your goal is to qualify them using BANT (Budget, Authority, Need, Timeline). Ask conversationally — don't interrogate. Start by confirming they signed up, then naturally explore: - What problem they're trying to solve (Need) - Their rough budget range (Budget) - Whether they're the decision maker (Authority) - When they need a solution (Timeline) If all four criteria are met, offer to book a demo with our senior team. If they're not ready, ask when a good time to follow up would be. Keep every response under 2 sentences. Be professional but warm. Don't be pushy — if they say they're busy, offer to call back.
Enable tools: CRM Lookup, Update CRM, Create Deal, Book Appointment, Schedule Callback, Send SMS.
Step 2: Create the Campaign
Step 1 of wizard: What & Who
- Name: "Inbound Lead Qualification - August"
- AI Objective: "Qualify inbound leads using BANT criteria. Book demos for qualified leads."
- Opening: "Hi {{name}}, this is Alex from [Company]. You recently signed up on our website — I wanted to check if you have a couple of minutes to chat about what you're looking for?"
- Qualifying criteria: "Budget above $5k", "Decision maker or can connect to one", "Need identified", "Timeline within 6 months"
- Interest keywords: "yes, book it, let's schedule, sounds great, interested"
Step 2: AI & Handler + CRM
- Agent: Select your qualification agent
- Pipeline: Your sales pipeline → Stage: "Qualified"
- Interested → SQL lifecycle
- Fields to collect: Budget, Timeline, Decision Maker, Company Size, Current Solution
- Channel: Voice | Number: your outbound number
Step 3: Delivery
- Schedule: 10:00–18:00, Mon–Fri
- Concurrent: 3 calls
- Max attempts: 3
- Pacing: Adaptive
- Voicemail: AMD enabled, action = "drop message"
- Goal: Stop at 50 interested (or whatever your monthly target is)
Step 3: Set Up Campaign Chaining
For contacts who are undecided (score 40–70), automatically feed them into a nurture campaign that runs 2 weeks later with a different approach:
- Create a second campaign: "Lead Nurture - Follow Up"
- Different AI objective: "You spoke with this person 2 weeks ago. They were interested but not ready. Check if anything has changed. Share a case study or offer."
- On your main campaign, set Chain Config → On Undecided → Campaign: "Lead Nurture" → Delay: 14 days
Step 4: Set Up Speed-to-Lead (Optional)
For maximum conversion, dial new leads within seconds of form submission:
- In your campaign, go to the API/Trigger section
- Get the trigger URL:
POST /api/v1/campaigns/:id/trigger - Connect your form/CRM to send a webhook when a new lead comes in:
Webhook payload (from your form/CRM)
{ "phone_number": "+919876543210", "name": "Rahul Kumar", "action": "instant_dial" }
The contact is added with highest priority and dialed within seconds — not waiting for the next batch cycle.
Step 5: Monitor & Iterate
During the campaign
- Watch the health score — if it drops below 50, your script may need adjustment
- Check engagement scores: if most contacts are 40–55, the pitch isn't landing — try a different angle
- If answer rate is low (<20%), try Local Presence caller ID or adjust schedule hours
After completion
- Read the AI Report → Key findings section tells you what patterns emerged
- Check Objections: what did people push back on? Adjust the script.
- Look at score distribution: healthy campaigns have a bell curve centered around 50–60, not clustered at 0–20.
Key Metrics to Track
Answer Rate
Good: > 30%
Bad: < 15%
Fix: Try Local Presence caller ID or different hours
Interest Rate
Good: > 15%
Bad: < 5%
Fix: Rework AI objective and opening message
Avg Score
Good: > 50
Bad: < 35
Fix: Your audience may not be ready. Try warmer leads.
Avg Duration
Good: > 90s
Bad: < 30s
Fix: People are hanging up fast. Softer opening needed.
