Inbound Call Handling (Answering Agent)

Your AI-powered answering agent handles incoming calls 24/7—answering questions, qualifying leads, quoting prices, and booking appointments directly into your calendar.

How It Works

When a customer calls your business (via call forwarding or direct number):

1. AI answers with your custom greeting
2. Identifies service need - Understands what the customer wants
3. Answers questions - Quotes pricing from your knowledge base
4. Collects information - Name, phone, address (only what’s needed)
5. Books appointment - Checks calendar and schedules in real-time
6. Confirms details - Reads back appointment time and service
7. Ends call naturally - “Anything else I can help with?” then goodbye

Average call duration: 2-4 minutes
Success rate: 85%+ of calls result in appointment or detailed lead capture

Call Flow Breakdown

1. Greeting & Opening

The AI answers with your configured greeting:

Example greeting:

“Thanks for calling ABC Plumbing! This is Sarah, how can I help you today?”

What happens:

• Uses your business name
• Introduces itself by agent name
• Opens with a clear question

2. Understanding the Service Need

The customer states their reason for calling:

Customer examples:

• “My garage door won’t open”
• “I need a plumber for a leaking faucet”
• “Do you do HVAC maintenance?”

What the AI does:

• Listens for service keywords
• Matches to your service catalog (from Custom Rules or Knowledge Base)
• Does NOT ask diagnostic questions - e.g., won’t ask “What brand is your garage door?”
• Simply acknowledges: “Got it, I can help you with that.”

If service is unclear:

• AI asks: “What service do you need help with?”
• Does NOT provide examples - avoids saying “Are you looking for repair or installation?“

3. Handling Pricing Questions

If the customer asks “How much will this cost?”:

AI searches your knowledge base for pricing information:

If pricing exists:

• Quotes the price directly
• Example: “A garage door spring replacement is $200.”

If multiple service variations exist:

• AI asks ONE clarifying question
• Example: “Is it a single or double garage door?”
• Then quotes the specific price

If pricing is complex or unavailable:

• AI offers on-site estimate (if you enabled “Free on-site estimates”)
• Example: “We offer free on-site estimates. The technician will give you exact pricing when they arrive. Would you like to schedule that?”

Then immediately proceeds to booking:

• “Let’s get you scheduled.”

Pricing Configuration

To enable pricing:

• ✅ Turn on “Pricing Enabled” in agent settings
• Upload pricing documents to Knowledge Base:
  • PDF price list
  • Text file with services and costs
  • Paste pricing into Custom Rules
• Enable “Free on-site estimates” if applicable

Example Knowledge Base entry:

Garage Door Services Pricing:
- Spring Replacement (Single Door): $200
- Spring Replacement (Double Door): $250
- Roller Replacement: $150
- Full Tune-Up: $99
- Emergency Service (after hours): $199 service call fee

4. Information Collection

The AI collects only the information necessary to book the appointment or take a message.

Standard fields (for appointment booking):

• ✅ Name
• ✅ Phone number
• ✅ Service address

Optional fields:

• ☐ Email
• ☐ Zip code
• ☐ Preferred date/time

How collection works:

1. Checks if customer already provided it - If they said “I’m John at 123 Main St,” the AI won’t ask again
2. Asks one question at a time - Never bundles questions together
3. Confirms phone number once - Reads it back to verify: “Got it, 909-321-9793.”

What the AI will NOT ask (unless you explicitly configure it):

• ❌ Email address (unless enabled)
• ❌ “Best time to call back”
• ❌ “How did you hear about us?“

5. Appointment Scheduling

Once all information is collected, the AI immediately checks availability.

No preamble - The AI doesn’t say:

• ❌ “Let me check the calendar for you”
• ❌ “Would you like me to book an appointment?”

It just invokes the scheduling tool automatically.

Presenting availability:

“Earliest I have is Monday 9 to 11 AM, or Wednesday 2 to 4 PM. Which works better for you?”

Best practices the AI follows:

• Always includes AM/PM with times
• Says “today” or “tomorrow” if applicable
• Rounds times (e.g., “2:00 PM” not “1:59 PM”)
• Offers 2-3 options

If customer needs different times:

• AI checks again with adjusted date range
• Can check specific dates: “Okay, let me check Thursday…”

When customer selects a time:

• AI books it immediately
• Confirms: “Perfect! You’re all set for…“

6. Appointment Confirmation

After booking, the AI reads back key details:

Default confirmation:

“Perfect! I’ve got you scheduled for Monday, March 15th from 9 to 11 AM for garage door spring replacement. You’ll receive a confirmation text shortly.”

Custom confirmation templates: You can configure exactly what the AI says. Example:

“All set! [TECH NAME] will be at [ADDRESS] on [DATE] between [TIME] for [SERVICE]. Watch for a confirmation text.”

7. Handling Edge Cases

Customer Wants to Add/Change Service

If customer booked, then says: “Actually, can you also check my garage door opener?”

AI response:

• Says: “Got it, I’ll add that to the notes for the technician.”
• Adds note to the appointment
• Does NOT book a second appointment

Returning Call / Callback Request

If customer says: “I got a message to call back” or “Someone called me from this number”

AI response:

• “Got it, I’ll let them know you called back.”
• Immediately follows Note Taking Policy (name + phone only)
• Does NOT ask “Can I try to help?”

Status Updates / Existing Projects

If customer calls about: “I want to check on my estimate” or “When is the technician arriving?”

AI response:

• Follows Transfer Policy or Note Taking Policy
• Does NOT try to schedule new work

Emergency Requests

If customer says: “This is an emergency” or “I need someone ASAP”

AI response:

• “I understand this is urgent. Earliest I have is [TIME]. Can I book that and mark it as urgent so the team can try to fit you in sooner?”
• Adds “EMERGENCY” to appointment notes

8. Call Transfer (If Enabled)

If call transfer is enabled, the AI will transfer calls that meet criteria:

Transfer criteria:

1. Customer explicitly demands human/manager AND refuses AI help
2. Booking fails after multiple attempts
3. Request is outside AI capabilities AND seems urgent

Transfer process:

Warm Transfer (Recommended):

1. AI says to customer: “Let me connect you with someone who can help.”
2. AI puts customer on hold
3. AI calls you and explains: “I have John on the line about garage door repair. Ready?”
4. You say “yes” → AI connects the call

Cold Transfer:

1. AI says: “Let me transfer you.”
2. Directly connects to your number

If transfer fails:

• AI says: “The transfer didn’t go through. Can I take a message and have someone call you back?”
• Follows Note Taking Policy

9. Note Taking Policy (Fallback)

If the AI cannot book an appointment (customer declines, calendar unavailable, transfer fails), it takes a detailed message.

What the AI collects:

• ✅ Name
• ✅ Phone number
• ✅ Brief description of what customer needs

What the AI says:

“I’ll make sure they get the message and someone will call you back shortly.”

What the AI does NOT do:

• ❌ Ask for address (not needed for callback)
• ❌ Ask for service details beyond what customer already stated
• ❌ Try to book an appointment after customer declined

10. Call Closing

Before ending the call:

If “Ask anything else” is enabled:

• AI asks: “Anything else I can help with?”
• Waits for response
• If yes → Addresses the question
• If no → Ends call

Ending the call:

• AI says an appropriate goodbye:
  • “Great! See you Monday.”
  • “You’re all set. Have a great day!”
  • “Talk to you soon!”
• AI invokes end_call function - does NOT just stop talking

Special Scenarios

After-Hours Calls

If you configure business hours, the AI can adjust behavior:

After-hours greeting example:

“Thanks for calling ABC Plumbing! We’re currently closed, but I can help you schedule an appointment or take a message.”

After-hours booking:

• AI still books appointments for future dates
• Can mark urgent requests as “EMERGENCY”

Voicemail Detection

If the AI calls someone and gets voicemail:

• AI detects answering machine messages
• Waits for the beep
• Leaves a configured voicemail message
• Marks call as “voicemail” in call log

Conversation Guidelines

The AI follows these natural conversation rules:

One Question at a Time

❌ “What’s your name and phone number?”
✅ “What’s your name?” → [waits for answer] → “And your phone number?”

Short Responses (10-12 words max)

❌ “Thank you so much for calling ABC Plumbing today. I really appreciate you reaching out to us and I’m here to assist you with whatever you need.”
✅ “Thanks for calling ABC Plumbing. How can I help today?”

Compassionate Responses

❌ Customer: “My water heater burst.” AI: “Great! Let’s get that scheduled.”
✅ Customer: “My water heater burst.” AI: “I understand, let’s get someone out there quickly.”

No Examples When Asking Questions

❌ “What’s the service address? For example, 123 Main Street.”
✅ “What’s the service address?”

Uses Information Already Provided

❌ Customer said their name is John. AI later asks: “What’s your name?”
✅ AI uses “John” in later conversation without re-asking

Monitoring & Optimization

Call Logs Dashboard

After each call, view:

• Full transcript - Read exactly what was said
• Call recording - Listen to the audio
• Extracted data - Name, phone, service, appointment details
• Call analytics:
  • Call duration
  • Customer sentiment (positive, neutral, negative)
  • Call success status (appointment booked, message taken, transferred)
  • User intent detected

Learn more about Call Logs & Analytics →

Improving Performance

Based on call logs, you can:

1. Refine greetings - Make them more natural
2. Add to knowledge base - Upload FAQs customers ask
3. Adjust pricing - Update service costs
4. Customize rules - Handle edge cases better

Configuration Options

Essential Settings

Agent Name
What the AI introduces itself as (e.g., “Sarah”, “your assistant”)

Company Name
Your business name as stated on calls

Custom Greeting
First thing the AI says when answering

Voice Selection
Choose from 50+ AI voices (male/female, various accents)

Service Identification

Enable Service Identification (Recommended: ON)
Allows AI to understand and match customer requests to your services

Service Question
How AI asks about services: “What service do you need?” or “How can I help?”

Custom Rules
Service-specific logic:

• “For spring replacement, ask if it’s single or double door”
• “We don’t service brand X garage doors”
• “Emergency calls should be marked URGENT”

Pricing

Enable Pricing (Recommended: ON)
AI quotes prices from knowledge base

Free On-Site Estimates (Optional)
AI offers free estimates when pricing is unavailable

Summarize Service Components (Optional)
AI breaks down pricing: “That’s $200—$150 for parts and $50 for labor”

Ask About Part Supply (Optional)
For repair businesses: “Do you have the part or do you need us to supply it?”

Appointment Scheduling

Enable Appointment Scheduling (Recommended: ON if you have calendar integration)

Scheduling Service:

• Cal.com
• FieldPulse
• GoHighLevel

Collection Fields:

• Name (required)
• Phone (required)
• Service Address (required)
• Email (optional)
• Zip Code (optional)

Appointment Summary Template
Customize confirmation message

Call Handling

Ask Anything Else Before Closing (Recommended: ON)
AI asks if customer needs anything else before ending call

Call Transfer Enabled (Optional)
Allow AI to transfer complex calls to you

Transfer Phone Number
Your number for transfers

Transfer Type:

• Warm Transfer - AI explains situation before connecting
• Cold Transfer - Direct connection

Private Handoff Prompt
What AI says to YOU during warm transfer

Best Practices

✅ Do This

• Test your agent weekly with real calls
• Review call logs regularly (at least weekly)
• Update knowledge base when customers ask new questions
• Keep greetings short and friendly
• Enable pricing if you have standard service costs
• Use warm transfers (not cold) for better customer experience

❌ Avoid This

• Don’t make greetings too long (over 20 words)
• Don’t add services to knowledge base that you don’t offer
• Don’t expect AI to diagnose technical issues
• Don’t configure too many custom rules (keep it simple at first)
• Don’t skip testing after making changes

Next Steps

1. Set up call forwarding - Call Forwarding Guide →
2. Configure your agent - Agent Configuration →
3. Add knowledge base content - Knowledge Base Guide →
4. Monitor call performance - Call Logs & Analytics →

Questions? Contact support at support@nzleads.com or use the in-app chat.