# CDP Domain Knowledge

> Understand Customer Data Platforms covering data collection, identity resolution, segmentation, and marketing activation

- Skill: `danhvb/cdp-domain-knowledge` (Agent Skill)
- Install (CLI): `npx skillmds@latest add danhvb/cdp-domain-knowledge`
- Raw SKILL.md: https://api.skillmd.com/api/skills/danhvb/cdp-domain-knowledge/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Marketing & Growth
- Author: danhvb (https://skillmd.com/u/danhvb)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/danhvb/cdp-domain-knowledge

---


# CDP Domain Knowledge Skill

## Purpose
Provide comprehensive CDP domain knowledge for analyzing requirements in customer data unification, segmentation, and marketing activation.

## What is a CDP?

A **Customer Data Platform (CDP)** is a packaged software that:
- Collects customer data from multiple sources
- Creates unified customer profiles (360-degree view)
- Enables segmentation and audience building
- Activates data to marketing channels

### CDP vs DMP vs CRM

| Capability | CDP | DMP | CRM |
|------------|-----|-----|-----|
| Data Type | First-party + Third-party | Third-party cookies | First-party |
| Identity | Known + Anonymous | Mostly anonymous | Known customers |
| Data Persistence | Long-term | Short-term (90 days) | Long-term |
| Primary Use | Personalization, Analytics | Advertising | Sales/Service |
| User Management | Individuals | Segments/Audiences | Accounts/Contacts |

## CDP Architecture

### Data Flow
```
Data Sources → Ingestion → Identity Resolution → Profile Unification → 
Segmentation → Audience Building → Channel Activation
```

### Core Components

**1. Data Collection**:
- Website/app events (page views, clicks, transactions)
- CRM data (contacts, accounts, opportunities)
- Transaction data (orders, returns, support tickets)
- Marketing data (email opens, campaign responses)
- Offline data (in-store purchases, call center)
- Third-party data (demographics, intent)

**2. Identity Resolution**:
- Deterministic matching (email, phone, customer ID)
- Probabilistic matching (device fingerprinting, behavior)
- Cross-device stitching
- Household linkage
- Anonymous to known conversion

**3. Profile Unification**:
- Golden record creation
- Attribute standardization
- Conflict resolution (which source wins?)
- Profile enrichment
- Historical event storage

**4. Segmentation**:
- Rule-based segments
- AI/ML-based segments
- Dynamic vs. static segments
- Lookalike audiences
- Predictive segments

**5. Activation**:
- Real-time personalization
- Email/SMS marketing
- Advertising platforms
- Customer service systems
- Analytics/BI tools

## Key CDP Concepts

### Identity Resolution

**Matching Keys**:
- Email address (primary)
- Phone number (hashed)
- Customer ID / Loyalty ID
- Device IDs (IDFA, GAID, cookie)
- Address + Name combination

**Identity Graph**:
```
            Cookie_A
               ↓
Email_1 ← Customer_123 → Device_B
               ↓
          Phone_555
```

**Confidence Levels**:
- High: Email match, customer ID
- Medium: Phone + first name
- Low: Cookie + behavior patterns

### Customer Profile

**Profile Attributes**:
```markdown
## John Smith - Customer Profile

### Identity
- Customer ID: CUST-123456
- Email: john.smith@email.com
- Phone: +1-555-123-4567
- Device IDs: [device_a, device_b]

### Demographics
- Age: 35
- Gender: Male
- Location: San Francisco, CA
- Household Income: $100-150K

### Behavioral
- First Purchase: 2024-01-15
- Last Purchase: 2026-01-10
- Total Orders: 12
- LTV: $1,245
- Average Order Value: $103.75
- Favorite Category: Electronics
- Last Page Viewed: Laptops
- Email Engagement: High (45% open rate)

### Segments
- VIP Customer
- Tech Enthusiast
- High Email Engager
- Likely to Churn (ML score: 0.65)

### Consent
- Email Marketing: Opted In
- SMS: Opted Out
- Third-party Sharing: Opted Out
- Last Updated: 2026-01-01
```

### Segmentation

**Segment Types**:

**Static Segments**: Fixed list, doesn't update automatically
```
Customers who purchased in Black Friday 2025
```

**Dynamic Segments**: Updates as data changes
```
Customers who purchased in last 30 days
AND total spend > $500
AND category = "Electronics"
```

**Predictive Segments**: ML-based predictions
```
Customers with churn risk > 70%
Customers likely to purchase in next 7 days
```

**Lookalike Segments**: Similar to seed audience
```
Customers similar to "VIP Customers" segment
```

### Data Activation

**Activation Channels**:
- **Email**: Personalized campaigns, triggered emails
- **SMS/Push**: Time-sensitive notifications
- **Advertising**: Facebook, Google, programmatic
- **Website**: Real-time personalization
- **Customer Service**: Agent context
- **Analytics**: Reporting and insights

**Real-time vs Batch**:
- Real-time: Website personalization, triggered emails
- Batch: Daily email campaigns, weekly reports

## Privacy & Consent

### Consent Management

**Consent Types**:
- Marketing email opt-in/out
- SMS opt-in/out
- Third-party data sharing
- Cookie consent
- Cross-site tracking

**Consent Requirements**:
- GDPR (EU): Explicit opt-in required
- CCPA (California): Opt-out right
- LGPD (Brazil): Consent required
- PDPA (Thailand/Singapore): Consent required

### Data Rights

**User Rights (GDPR/CCPA)**:
- Right to access (what data do you have?)
- Right to rectification (fix incorrect data)
- Right to erasure (delete my data)
- Right to portability (give me my data)
- Right to opt-out (don't sell my data)

## Key Metrics

### Data Quality
- Profile completeness (% fields populated)
- Identity match rate
- Data freshness (recency of updates)
- Duplicate rate

### Engagement
- Segment size and growth
- Activation success rate
- Channel reach
- Personalization lift

### Business Impact
- Revenue influenced by CDP
- Customer lifetime value improvement
- Campaign ROI improvement
- Reduced customer acquisition cost

## Common CDP Requirements

### Data Collection
```
REQ-001: System shall collect website events including page views, product views, add-to-cart, and purchases
REQ-002: System shall ingest CRM data via API daily
REQ-003: System shall capture email engagement events in real-time
```

### Identity Resolution
```
REQ-010: System shall match profiles using email as primary key
REQ-011: System shall stitch anonymous cookie to known profile on login
REQ-012: System shall maintain identity graph with confidence scores
```

### Segmentation
```
REQ-020: System shall support dynamic segment creation using AND/OR logic
REQ-021: System shall recalculate segment membership daily at minimum
REQ-022: System shall support audience size estimation before creation
```

### Activation
```
REQ-030: System shall sync audiences to Facebook in near real-time
REQ-031: System shall trigger email via Klaviyo when user enters segment
REQ-032: System shall provide personalization API for website with < 100ms latency
```

## Questions for Stakeholders

- What data sources need to be connected?
- How do you identify customers today?
- What segments do you need?
- Which marketing channels for activation?
- What personalization use cases?
- Privacy/consent requirements?
- Real-time vs batch needs?

## Popular CDP Platforms

- **Segment**: Developer-friendly, data infrastructure
- **mParticle**: Mobile-focused, SDK expertise
- **Treasure Data**: Enterprise, data lake integration
- **Tealium**: Tag management heritage
- **Adobe Experience Platform**: Adobe ecosystem
- **Salesforce CDP**: Salesforce ecosystem

## References

- CDP Institute Resources
- First-party Data Strategy
- Privacy Regulations (GDPR, CCPA)