🔗 Identity Stitching in Adobe Experience Platform

            The Big Picture: Identity stitching is how AEP connects the dots between different identifiers (cookies, emails, phone numbers) to realize they all belong to the same person, creating one unified customer profile.
        

🎯 The Challenge: Why Identity Stitching is Needed

❌ Without Identity Stitching

Sarah appears as 4 different people:

📱 Mobile Cookie: anonymous_user_123
💻 Desktop Cookie: visitor_xyz_789
📧 Email: sarah@gmail.com
🛍️ Loyalty ID: GOLD_12345

Result: Fragmented data, poor personalization, duplicate marketing

✅ With Identity Stitching

Sarah is recognized as ONE person:

👤 Complete Sarah Profile

• All browsing history (mobile + desktop)

• Email engagement

• Purchase history

• Store visits

• Loyalty status

Result: Unified view, perfect personalization, relevant marketing

🔍 Step-by-Step: How Identity Stitching Actually Works

📖 Real-World Example: Sarah's Journey

Let's follow Sarah through her customer journey and see exactly how AEP connects her identities:

1 📱 Monday: Sarah browses on mobile (Anonymous)

What happens:

Web SDK creates new ECID: ECID_ABC123
Tracks: viewed homepage, browsed shoes, 5 minutes on site
AEP creates anonymous profile linked to ECID_ABC123

Profile in AEP:
{
  "identities": ["ECID_ABC123"],
  "person": "unknown",
  "events": ["homepage_view", "product_browse_shoes"]
}
                

2 💻 Tuesday: Sarah browses on desktop (Still Anonymous)

What happens:

Different device = New ECID: ECID_XYZ789
Tracks: viewed dresses, added item to cart
AEP creates SECOND anonymous profile (doesn't know it's same person yet)

Profile #1: ECID_ABC123 (mobile)
Profile #2: ECID_XYZ789 (desktop) - SEPARATE!
                

3 📧 Wednesday: Sarah signs up for newsletter on desktop

THE MAGIC MOMENT - First Identity Link!

Sarah enters email: sarah@gmail.com
Web SDK sends both identities: ECID_XYZ789 + sarah@gmail.com
AEP Identity Service creates first connection!

🔗 Identity Graph Update

ECID_XYZ789 ↔ sarah@gmail.com

Now AEP knows desktop browsing belongs to sarah@gmail.com

4 📱 Thursday: Sarah logs into mobile app

THE SECOND CONNECTION!

Sarah logs in with: sarah@gmail.com
Mobile already has: ECID_ABC123
AEP connects: ECID_ABC123 ↔ sarah@gmail.com

🔗 Complete Identity Graph

ECID_ABC123 ↔ sarah@gmail.com ↔ ECID_XYZ789

BREAKTHROUGH: All three identities now linked!

5 🛍️ Friday: Sarah makes purchase in store

Adding offline identity:

Provides email at checkout: sarah@gmail.com
Gets loyalty card: LOYALTY_12345
POS system sends both to AEP

🔗 Final Complete Identity Graph

ECID_ABC123 ↔ sarah@gmail.com ↔ ECID_XYZ789 ↔ LOYALTY_12345

COMPLETE PICTURE: Mobile + Desktop + Email + Store = ONE CUSTOMER

🧠 How AEP Identity Service Works Behind the Scenes

The Technical Process

1. Identity Collection

Every time data comes into AEP, it includes identity information:

// Web SDK sends this data
{
  "identities": {
    "ECID": "ABC123",
    "Email": "sarah@gmail.com"
  },
  "eventData": {
    "productView": "running-shoes"
  }
}
            

2. Identity Graph Building

AEP maintains a graph database that looks like this:

Identity Type	Identity Value	Connected To	First Seen
ECID	ABC123	sarah@gmail.com	Monday 2:30 PM
Email	sarah@gmail.com	ABC123, XYZ789, LOYALTY_12345	Wednesday 4:15 PM
ECID	XYZ789	sarah@gmail.com	Tuesday 6:45 PM
Loyalty ID	LOYALTY_12345	sarah@gmail.com	Friday 11:20 AM

3. Profile Merging

When identities are linked, AEP merges all associated data:

👤 Sarah's Unified Profile After Stitching

Identities: ECID_ABC123, ECID_XYZ789, sarah@gmail.com, LOYALTY_12345

All Events:

Monday mobile: homepage view, shoe browsing
Tuesday desktop: dress viewing, cart addition
Wednesday desktop: newsletter signup
Thursday mobile: app login
Friday store: purchase, loyalty signup

Calculated Attributes:

Total sessions: 5
Favorite category: Footwear
Preferred device: Mobile (60%)
Customer value: High (made purchase)

🎯 Types of Identity Stitching

🎯 Deterministic Stitching

Based on exact matches

Email addresses
Phone numbers
Loyalty card numbers
Customer IDs

                    Accuracy: 100% certain

                    When: Customer provides identifying information

🤖 Probabilistic Stitching

Based on behavioral patterns & device fingerprinting

IP address patterns
Device characteristics
Browsing behavior
Location data

                    Accuracy: 80-95% confidence

                    When: Customer remains anonymous

💼 Business Impact & Interview Points

Before Identity Stitching

❌ Sarah gets irrelevant email about men's shoes (based only on mobile browsing)
❌ Website shows generic homepage on desktop (doesn't know her preferences)
❌ Marketing team thinks they have 4 different customers
❌ Customer service can't see full purchase history

After Identity Stitching

✅ Personalized email featuring dresses (knows she browsed them on desktop)
✅ Mobile app shows running shoes prominently (knows her interests)
✅ Consistent experience across all channels
✅ Complete customer service view
✅ Accurate analytics and reporting

🚀 Key Interview Talking Points

Core Explanation:

"Identity stitching is how AEP recognizes that different digital footprints belong to the same person. It starts with cookies tracking anonymous behavior, then connects those cookies to identifying information like email addresses when customers log in or make purchases. This creates a complete, unified customer profile that enables true personalization."

Business Value Statement: "Without identity stitching, we'd be shooting in the dark with personalization. With it, we can deliver consistent, relevant experiences whether a customer is browsing on mobile, shopping on desktop, or visiting our store."

Interview Q&A:

Q: "How does identity stitching handle privacy?"
A: "AEP respects customer consent and privacy settings. If a customer opts out, their identities won't be stitched, and we maintain separate anonymous profiles. The system is built to comply with GDPR and CCPA requirements."

Q: "What happens if identity stitching makes a mistake?"
A: "AEP has governance controls and confidence thresholds. Deterministic matches (email, phone) are 100% accurate. Probabilistic matches have confidence scores, and businesses can set minimum thresholds. There are also manual review processes for edge cases."
        

🔧 Common Implementation Scenarios

E-commerce Site

Cookie → Email (newsletter) → Loyalty ID (purchase) → Phone (customer service)

Banking App

Device ID → Customer ID (login) → Phone (verification) → SSN (account opening)

Media Site

Anonymous viewing → Email (subscription) → Social login → Mobile app ID

            Remember: Identity stitching is the foundation that makes everything else in AEP possible - segmentation, personalization, journey orchestration, and analytics all depend on having unified customer profiles!
        