Lead Qualification Agent
In this tutorial, you’ll build a voice agent that qualifies leads through natural conversation. It uses the BANT framework (Budget, Authority, Need, Timeline) to assess fit, scores each lead, logs the data to your CRM, and routes qualified leads to your sales team.What You’ll Build
A phone agent that:- Engages leads in a natural, consultative conversation
- Qualifies them using the BANT framework without sounding like an interrogation
- Scores leads as Hot, Warm, or Cold
- Logs all lead data to your CRM via webhook
- Records responses in a Google Sheet for pipeline tracking
- Sends a follow-up SMS with next steps
- Routes hot leads to a human sales rep
Step 1: Create the Agent
- Go to your thinnestAI Dashboard
- Click Create Agent
- Name it:
Lead Qualification Agent - Set the description:
Qualifies inbound and outbound leads using BANT framework
Step 2: Choose the Model
Select Claude Sonnet as your model. Claude Sonnet excels at following nuanced conversation flows. It naturally weaves qualifying questions into the dialogue rather than reading from a script, which builds trust with prospects.Step 3: Assign Tools
CRM Webhook
This is the primary data capture tool. It sends qualified lead data to your CRM.- Go to the Tools tab
- Enable CRM Webhook
- Configure your endpoint:
SMS
Enable SMS to send follow-up information to leads after the call.- Enable SMS in the Tools tab
- Configure your sender number
Google Sheets
Log every lead to a spreadsheet for pipeline tracking and team visibility.- Enable Google Sheets in the Tools tab
- Connect your Google account
- Create a sheet with these columns:
Name | Company | Phone | Email | Role | Budget | Authority | Need | Timeline | Score | Date | Notes - Share the sheet URL with the agent
Step 4: Set the System Prompt
Copy and paste this entire prompt into the System Prompt field:Important: Replace all bracketed placeholders with your actual company details, product info, and benefits.
Step 5: Configure for Outbound (Optional)
If you want to use this agent for outbound calling campaigns:- Go to the Campaigns tab
- Create a new campaign
- Upload your lead list (CSV with Name, Phone, Company columns)
- Set the calling schedule (e.g., weekdays 9 AM–5 PM in the lead’s timezone)
- Set the call pace (e.g., 5 concurrent calls)
Step 6: Assign a Phone Number
- Go to the Phone tab
- Assign a phone number (use a local area code matching your market for higher answer rates)
- Set call direction to Inbound, Outbound, or Both
- Save
Step 7: Test Your Agent
Testing Checklist
| Scenario | What to Test | Expected Behavior |
|---|---|---|
| Hot lead | Express urgent need, confirm budget and authority | Agent scores as hot, offers to schedule demo |
| Warm lead | Show interest but vague on timeline and budget | Agent scores as warm, offers to send info and follow up |
| Cold lead | Say you’re just browsing, no budget | Agent scores as cold, politely wraps up, sends resources |
| Objection: competitor | Say you already use a competitor | Agent asks what they like/dislike, positions differentiators |
| Objection: price | Say it sounds expensive | Agent reframes around ROI and cost of current problem |
| ”Send me info” | Ask for info without engaging | Agent asks one qualifying question first |
| Outbound cold open | Answer an outbound call | Agent introduces themselves, references trigger event, asks permission |
| Not interested | Firmly say you’re not interested | Agent thanks them, doesn’t push, ends call gracefully |
What to Listen For
- Natural flow: Does it feel like a conversation or a survey?
- Question spacing: Are BANT questions spread out, not clustered?
- Active listening: Does the agent reference what the prospect said earlier?
- CRM data: After the call, check your CRM webhook — is the payload complete and accurate?
- Lead score accuracy: Does the score match the conversation?
Tips for Improvement
- A/B test opening lines. Try different outbound openers and track which ones get the highest engagement rates.
- Add industry-specific knowledge. Upload case studies and ROI data so the agent can cite relevant examples during objection handling.
- Set up lead routing. Configure your CRM webhook to auto-assign hot leads to sales reps and trigger a Slack notification.
- Track conversion rates. Monitor how many hot leads actually convert to demos and deals. Use this to refine your scoring criteria.
- Create segments. Build different campaigns for different industries or company sizes, each with tailored talking points in the system prompt.
- Follow-up sequences. Set up a campaign to call warm leads back after 2 weeks. Persistence (without being annoying) dramatically improves conversion.
Next: Order Management Agent — Build an agent that handles order inquiries, modifications, and returns.

