Enterprise Affinity & Engagement Architecture

Designed and implemented an enterprise-scale behavioral intelligence architecture within Salesforce Salesforce Marketing Cloud to transform fragmented campaign engagement into reusable, actionable first-party intelligence across 35+ brands.

The framework operationalized affinity capture, engagement enrichment, lifecycle segmentation, and scalable personalization infrastructure—creating a reusable CRM intelligence layer that powered both always-on automation and future downstream activation use cases.


Executive Summary

At Diageo North America, CRM data existed primarily as isolated campaign activity tied to individual sends and disconnected brand programs. Subscriber behavior was not being consistently translated into reusable intelligence.

I architected and implemented a scalable Affinity + Engagement data framework that standardized how behavioral signals were captured, stored, enriched, and operationalized across the SFMC ecosystem.

The solution introduced:

  • centralized affinity data structures
  • engagement intelligence tables
  • reusable segmentation architecture
  • automated nightly enrichment processes
  • lifecycle-aware behavioral tracking
  • modular personalization readiness

This transformed CRM from a channel used primarily for outbound campaigns into a continuously learning behavioral intelligence system.


The Business Problem

The existing CRM ecosystem faced several structural limitations:

Fragmented Behavioral Data

Engagement signals lived inside individual journeys or campaign reporting and were rarely reusable.

No Unified Intelligence Layer

Brands operated independently without a scalable framework for affinity capture or behavioral enrichment.

Personalization Was Operationally Expensive

Each audience or personalized send required manual segmentation logic and duplicated effort.

Lifecycle Journeys Could Not Learn

Journeys sent emails, but subscriber behavior was not systematically feeding future targeting or content decisions.

Scaling Across 35+ Brands Was Unsustainable

The organization needed a reusable architecture capable of supporting both enterprise governance and brand-level flexibility.


My Role

Lead CRM Architect / SFMC Product Lead

Owned:

  • enterprise data modeling strategy
  • SFMC architecture design
  • automation framework design
  • affinity taxonomy structure
  • segmentation logic
  • SQL enrichment processes
  • CloudPage behavioral capture strategy
  • operational deployment model
  • lifecycle intelligence framework

Worked cross-functionally with:

  • CDP teams
  • data engineering
  • CRM operations
  • activation teams
  • global platform stakeholders
  • external agencies

Architecture Overview

Core Framework Components

1. Welcome Entry Layer

Brand-level welcome tables captured acquisition source and entry metadata.

Examples:

  • {Brand}_Welcome
  • API-driven acquisition sources
  • multi-entry handling using composite keys

2. Master Sendable Layer

Standardized sendable records across brands.

Examples:

  • Master_Sendable_{Brand}

Contained:

  • subscriber identifiers
  • consent status
  • send eligibility
  • core profile attributes

3. Affinity Intelligence Layer

Centralized reusable behavioral indicators.

Examples:

  • {Brand}_Affinity

Tracked:

  • product preferences
  • flavor affinities
  • lifestyle interests
  • occasion signals
  • sports/music/culture interests
  • premiumization indicators
  • engagement-derived behavioral categories

Examples of affinity fields:

  • Affinity_Recipes
  • Affinity_Sports
  • Affinity_Hosting
  • Affinity_Events
  • Affinity_Futbol
  • Variant_Blanco
  • Variant_1942

4. Engagement Intelligence Layer

Persistent engagement scoring and behavioral activity tracking.

Examples:

  • {Brand}_Engagement

Tracked:

  • opens
  • clicks
  • recency
  • engagement velocity
  • behavioral enrichment
  • interaction timestamps
  • source attribution

5. Automation & Enrichment Layer

Nightly automation framework:

  • synchronized new entries
  • enriched affinity tables
  • updated engagement states
  • standardized segmentation logic
  • maintained reusable intelligence

Built using:

  • SQL automations
  • Journey Builder update activities
  • AMPscript data writes
  • API integrations
  • CloudPage submissions

Technical Strategy

Scalable Behavioral Modeling

The architecture intentionally separated:

  • sendable identity
  • behavioral intelligence
  • engagement activity

This prevented:

  • bloated sendable models
  • duplicated engagement logic
  • journey-specific silos

It also enabled:

  • reusable segmentation
  • modular personalization
  • scalable reporting
  • future CDP integration

Enterprise Personalization Readiness

The framework was designed to support:

  • dynamic content rotation
  • affinity-driven messaging
  • journey branching
  • behavioral suppression logic
  • audience modeling
  • paid media enrichment

Without requiring brands to rebuild segmentation logic repeatedly.


Operational Flexibility

The system balanced:

  • enterprise governance
  • local brand autonomy
  • reusable architecture
  • scalable deployment velocity

This allowed:

  • smaller brands to onboard quickly
  • large brands to support more complex personalization
  • CRM operations to scale without equivalent headcount growth

Outcomes & Impact

Enterprise Scale

Supported:

  • 35+ brands
  • millions of sends
  • reusable lifecycle infrastructure
  • ongoing affinity capture across campaigns and journeys

Net-New Behavioral Intelligence

Generated hundreds of thousands of reusable affinity signals through lifecycle engagement programs.

Created a persistent intelligence layer instead of one-time campaign reporting.


Operational Efficiency

Reduced manual segmentation overhead through:

  • reusable DE structures
  • standardized enrichment logic
  • modular automation frameworks

Enabled faster deployment timelines and reduced dependency on one-off builds.


Personalization Enablement

Created the foundational infrastructure for:

  • affinity-based personalization
  • dynamic content systems
  • behavioral audience segmentation
  • downstream CDP activation
  • future paid media lookalike modeling

Why This Matters

Most CRM organizations operate campaign systems.

This architecture transformed CRM into an intelligence system.

Instead of treating engagement as temporary reporting data, the framework operationalized behavioral signals into reusable enterprise assets that continuously improved targeting, personalization, and lifecycle orchestration over time.

That shift enabled CRM to function not simply as outbound marketing infrastructure—but as a scalable behavioral intelligence engine.