Repeat Caller Recognition

Instantly identify returning customers and provide personalized greetings to reduce handle time and improve satisfaction.

---

Business Problem

Scenario: Customer calls back about the same issue

Without context:

• Agent asks: "What can I help you with today?"
• Customer repeats entire story
• Agent searches for previous interaction
• Time wasted: 2-3 minutes

With Sticky Calls:

• System recognizes caller instantly
• Agent sees: "Customer called 2 days ago about billing issue"
• Agent: "Hi Sarah, I see you called about your billing question. Let me help you with that."
• Time saved: 2-3 minutes

ROI:

• ⏱️ 30% faster call resolution
• 📈 25% improvement in CSAT
• 💰 Reduced handle time = more calls handled
• 😊 Better customer experience

---

Implementation

1. Call Start - Identify Caller

curl -X POST https://api.stickycalls.com/v1/calls/start \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "call_id": "call_001",
    "identity_hints": {
      "ani": "+14155551234"
    }
  }'

Response (Returning Caller):

{
  "call_id": "call_001",
  "customer_ref": "a1b2c3d4-e5f6-7890-abcd-ef1234567890",
  "call_start": "2026-02-07T10:30:00.000Z",
  "identity": {
    "confidence": 0.95,
    "level": "very_high",
    "sources": ["ani:mobile", "recency:2day"],
    "recommendation": "reuse"
  },
  "open_intents": [],
  "variables": {
    "last_issue": {
      "value": "Duplicate billing charge - refund processed for $50",
      "source": null,
      "ttl_seconds": 2592000
    }
  }
}

2. Personalized IVR Greeting

Standard Greeting:

"Thank you for calling. Please hold while we connect you to the next available agent."

Personalized Greeting (High Confidence):

"Welcome back! I see you previously called about a billing issue. Let me connect you to our billing department."

Implementation:

IF identity.confidence >= 0.8:
    Play: "Welcome back, {customer_name}!"
    IF open_intents is empty:
        Play: "I see your previous issue was resolved. How can I help you today?"
    ELSE:
        Play: "I see you previously called about {open_intents[0].intent}. Let me connect you to the right team."
    Route to: Appropriate queue based on context
ELSE:
    Play: Standard greeting

3. Agent Screen Pop

Display context immediately when call routes:

┌─ Caller History ─────────────────────┐
│ Name: Sarah Johnson                   │
│ Phone: +1 (415) 555-1234              │
│                                       │
│ Previous Context:                     │
│ "Called 2 days ago about duplicate    │
│  billing charge. Refund processed     │
│  for $50. Customer was satisfied."    │
│                                       │
│ Last Call: Jan 26, 2025 2:30 PM      │
│ Confidence: 95%                       │
│ Open Issues: None                     │
└───────────────────────────────────────┘

4. Agent Script

Opening: "Hi Sarah, I see you called us recently about a billing issue. The notes say we processed a $50 refund for you. Is there anything else I can help you with today?"

Benefits:

• Customer feels remembered
• No need to repeat information
• Faster resolution
• Higher satisfaction

---

Example Scenarios

Scenario 1: Follow-Up Call

Day 1 - Initial Call:

• Customer: "I was charged twice for my subscription"
• Agent: Processes $50 refund
• Context saved: "Duplicate charge - refunded $50"

Day 3 - Follow-Up:

• Customer calls back
• System recognizes caller (confidence: 92%)
• IVR: "Welcome back! I see your refund was processed. Let me connect you to billing."
• Agent sees full history immediately
• Agent: "Hi! Just checking - did you receive the $50 refund?"
• Customer: "Yes! I just wanted to confirm it won't happen again."
• Agent: Quick resolution with no story repetition

Result:

• Handle time: 2 minutes (vs. 5 minutes without context)
• CSAT: 9/10 (vs. 7/10 typical)

Scenario 2: Recurring Technical Issue

Week 1:

• Customer: Internet keeps disconnecting
• Agent: Troubleshooting steps attempted
• Context: "Internet disconnection - rebooted modem, issue persists"

Week 2:

• Customer calls again
• System shows previous troubleshooting
• Agent: "I see we tried rebooting the modem last week. Let's try the next step..."
• No wasted time re-attempting same fixes

Scenario 3: VIP Customer

Last Month:

• Customer: Enterprise account inquiry
• Context saved: "VIP - Enterprise tier, $50K/year account"

This Month:

• Customer calls
• System identifies as VIP (confidence: 100%)
• Auto-route to: Priority queue
• Agent sees VIP status immediately
• Personalized service from start

---

Configuration Examples

Five9 IVR Script

<!-- Get caller context -->
<WebConnector name="StickyCallsStart" />

<!-- Branch based on match -->
<Branch>
  <Condition>matchFound = true AND confidence > 0.8</Condition>
  <Then>
    <PlayMessage>Welcome back! I see you previously called about $(context).</PlayMessage>
    <RouteToQueue>$(recommendedQueue)</RouteToQueue>
  </Then>
  <Else>
    <PlayMessage>Welcome to our contact center.</PlayMessage>
    <Menu>Standard IVR menu</Menu>
  </Else>
</Branch>

NICE CXone Studio

RESTful API Action: IdentifyCaller
  ↓
Branch on {matchFound} AND {confidenceScore} > 0.8
  ↓
TRUE:  Play "Welcome back, we have your information"
       Route to specialized queue
  ↓
FALSE: Standard greeting and menu

Generic IVR (Pseudocode)

# At call start
result = api.identify_caller(call_id, ani)

if result['match_found'] and result['confidence_score'] > 0.8:
    play_audio("welcome_back.wav")
    
    # Extract key info from context
    if "billing" in result['context'].lower():
        route_to_queue("billing_priority")
    elif "technical" in result['context'].lower():
        route_to_queue("technical_support")
    else:
        route_to_queue("general_priority")
        
    # Show context to agent
    screen_pop({
        "previous_context": result['context'],
        "last_call_date": result['last_call_at'],
        "confidence": result['confidence_score']
    })
else:
    play_audio("standard_greeting.wav")
    show_standard_menu()

---

Key Metrics

Track these KPIs to measure success:

Metric	Before	After	Improvement
Avg Handle Time (AHT)	5:30	3:45	-32%
First Call Resolution (FCR)	68%	85%	+25%
Customer Satisfaction (CSAT)	7.2/10	8.9/10	+24%
Repeat Call Rate	22%	12%	-45%
Agent Efficiency	8 calls/hr	11 calls/hr	+38%

---

Best Practices

1. Set Confidence Thresholds

> 0.9: Full personalization, auto-route
> 0.7: Personalization with verification
> 0.5: Show context to agent, standard flow
< 0.5: Treat as new caller

2. Save Rich Context

Bad:

"Issue resolved"

Good:

"Billing issue resolved. Customer reported duplicate charge for subscription ($50). Refund processed on 1/26. Customer confirmed satisfaction. No further action needed."

3. Update Context Throughout Call

Don't just save at the end - update as conversation progresses:

# During call
api.update_context(call_id, "Customer mentioned wanting to upgrade plan")

# At end
api.save_context(call_id, full_summary, open_intents)

4. Handle Low Confidence Gracefully

IF confidence 0.5-0.7:
    Agent sees context (for reference)
    BUT ask verification question:
    "Before we begin, can I confirm your account number?"

5. Train Agents

Agent Training Points:

• How to interpret confidence scores
• When to trust automated context
• How to verify customer identity if uncertain
• How to write good context summaries

---

ROI Calculator

Assumptions:

• 10,000 inbound calls/month
• 30% are repeat callers (3,000 calls)
• 2 minutes saved per repeat caller
• Agent cost: $25/hour

Time Savings:

• 3,000 calls × 2 minutes = 6,000 minutes = 100 hours/month
• Annual: 1,200 hours

Cost Savings:

• 1,200 hours × $25 = $30,000/year

Sticky Calls Investment: View current pricing →

Massive ROI - Save thousands on handle time while improving customer satisfaction

---

Implementation Checklist

• Sign up for Sticky Calls API at stickycalls.com
• Get API key from dashboard
• Integrate /calls/start at beginning of call
• Parse response and set IVR variables
• Create personalized greeting for high-confidence matches
• Configure agent screen pop with context display
• Integrate /calls/end to save context
• Test with known phone numbers
• Train agents on new workflow
• Monitor metrics (AHT, CSAT, FCR)
• Iterate and optimize

---

Next Steps

• View API Reference - Complete endpoint documentation
• Five9 Integration - Platform-specific guide
• NICE CXone Integration - Platform-specific guide
• Proactive Callback Use Case - Another scenario
• Contact Support - Get help

---

Ready to reduce handle time and improve CSAT? Sign up for free →