Salesforce Marketing Cloud Architecture Guide

Salesforce Marketing Cloud Architecture: Enterprise Guide

• Marketing Cloud Engagement: built for B2C and high-volume B2B — millions of contacts, email, SMS, push, and advertising.
• Account Engagement (formerly Pardot): built for B2B lead scoring and nurture with tight Sales Cloud integration.
• Three scripting languages: AMPscript (inline personalization), Server-Side JavaScript / SSJS (complex logic in CloudPages), and SQL (data queries in Automation Studio).
• Basic edition: approximately $400/month for 10,000 contacts. Corporate and Enterprise: $3,750–$65,000+/month.
• MuleSoft integration: $35,000+/year for complex Dynamics 365 or Microsoft data connections.
• EPC Group: 29 years of Microsoft consulting. Architects Marketing Cloud integrations with Dynamics 365, Azure, and Power BI.

Marketing Cloud Architecture Overview

Unlike Salesforce Sales Cloud and Service Cloud, Marketing Cloud runs on a separate technology stack. Salesforce acquired it from ExactTarget in 2013. This has significant architectural implications.

Core architectural characteristics:

• Separate data model — Marketing Cloud uses Data Extensions (its own storage) rather than Salesforce objects. Data is synced between Marketing Cloud and Sales Cloud via Marketing Cloud Connect.
• Studio-based architecture — functionality is organized into specialized studios: Email Studio, Mobile Studio, Social Studio, Advertising Studio, Journey Builder, Automation Studio, and Content Builder
• Multi-tenant cloud — runs on dedicated infrastructure with tenant-level isolation; Enterprise editions support multiple Business Units for organizational separation
• API-first design — REST and SOAP APIs are exposed for nearly every operation, enabling deep integration with external systems

The Data Model: Data Extensions Explained

The most critical architectural concept in Marketing Cloud is the Data Extension. Unlike Salesforce CRM objects, Data Extensions are relational tables that you design from scratch.

Types of Data Extensions:

• Sendable Data Extensions — contain subscriber data (email address or subscriber key); used as the audience for email sends and journey entries
• Non-Sendable Data Extensions — reference tables for lookups, personalization data, and reporting (product catalogs, store locations, preference data)
• Data Relationships — link Data Extensions to each other for relational personalization (e.g., link Subscribers to Orders to personalize emails with recent purchase data)
• Synchronized Data Extensions — automatically replicated from Salesforce CRM objects via Marketing Cloud Connect; read-only in Marketing Cloud
• Retention Policies — configure automatic data deletion based on age for GDPR, CCPA, and data retention compliance

Proper Data Extension design is critical. Poorly designed data models lead to slow query performance, personalization errors, and compliance gaps. EPC Group has seen implementations fail because of foundational data model mistakes.

Marketing Cloud Scripting Languages

Marketing Cloud uses three proprietary scripting languages. Each serves a different purpose:

• AMPscript — inline personalization in emails and landing pages; the primary tool for dynamic content insertion
• Server-Side JavaScript (SSJS) — complex logic in CloudPages and Script Activities; supports server-side rendering and external API calls
• SQL — data queries in Automation Studio for audience segmentation, data transformation, and Data Extension management

Marketing Cloud also exposes REST and SOAP APIs for external integrations. EPC Group's developers are proficient in all three scripting languages. Common external integrations use Node.js, Python, and C#/.NET.

Integration with Microsoft Dynamics 365

Organizations running both Marketing Cloud and Dynamics 365 need careful integration architecture. Options include:

• Marketing Cloud Connect for Dynamics — synchronizes Dynamics 365 contacts, leads, and campaign data with Marketing Cloud Data Extensions. Supports bidirectional sync for campaign responses.
• MuleSoft integration — for complex scenarios requiring custom data transformations; provides pre-built connectors for both platforms
• Azure Data Factory — for batch data movement between Dynamics 365 Dataverse and Marketing Cloud Data Extensions via scheduled ETL pipelines
• Power BI reporting — extract Marketing Cloud campaign data (opens, clicks, conversions, bounces) into Power BI for unified reporting alongside Dynamics 365 sales and service data
• Custom API integration — Marketing Cloud REST APIs enable real-time triggers from Dynamics 365 Power Automate workflows to initiate journeys or trigger transactional messages

Architecture Best Practices

1. Design your data model first — before building any campaigns or journeys, map all Data Extensions, relationships, and data flows. This is the foundation everything else depends on.
2. Use Business Units for separation — multi-division or multi-brand organizations should use Business Units to segregate data, content, and user access while maintaining shared assets.
3. Implement a subscriber key strategy — use a consistent, immutable subscriber key (not email address) across all channels. This prevents duplicate records when email addresses change.
4. Use SQL queries in Automation Studio — complex audience segmentation should use SQL rather than drag-and-drop filters for better performance on large-scale operations.
5. Plan for GDPR and privacy compliance from day one — implement consent management, preference centers, data retention policies, and audit trails at the start. Retrofitting compliance into an existing deployment costs far more.
6. Monitor deliverability proactively — configure dedicated sending domains, authenticate with SPF/DKIM/DMARC, implement sunset policies for inactive subscribers, and monitor deliverability metrics continuously.

Why EPC Group for Marketing Cloud Architecture

EPC Group has 29 years of enterprise consulting spanning both Microsoft and Salesforce ecosystems. We design Marketing Cloud architectures that integrate with Dynamics 365, Azure, and Power BI to create unified marketing and sales operations.

Our strengths in Marketing Cloud architecture:

• Cross-platform integration — Marketing Cloud plus Dynamics 365, Azure, and Power BI in one unified ecosystem
• Data architecture at scale — Data Extension models that perform for millions of contacts without slow queries or personalization errors
• Compliance-first approach — HIPAA for healthcare, SOC 2 for financial services, and GDPR for European operations built in from the ground up
• Journey optimization — multi-channel customer journeys across email, SMS, push, web, and advertising with AI-powered optimization
• Deliverability management — inbox placement rates above 95% through dedicated sending domains and proactive reputation management

Frequently Asked Questions

What is the difference between Marketing Cloud Engagement and Account Engagement?

Marketing Cloud Engagement (formerly ExactTarget) is designed for B2C and high-volume B2B marketing — millions of contacts, multi-channel campaigns, and advanced data management via Data Extensions.

Account Engagement (formerly Pardot) is for B2B lead scoring, nurture programs, and tight Sales Cloud integration. Most B2C and high-volume B2B organizations need Marketing Cloud Engagement. Smaller B2B teams often start with Account Engagement.

How much does Salesforce Marketing Cloud cost?

Pricing is based on edition and contact/message volume. Basic edition: approximately $400/month for 10,000 contacts. Corporate and Enterprise editions with Journey Builder, Automation Studio, and advanced features: $3,750 to $65,000+/month depending on volume and channels.

MuleSoft integration adds $35,000+/year. EPC Group helps clients optimize licensing to minimize cost while meeting all requirements.

Can Marketing Cloud integrate with Microsoft Dynamics 365?

Yes, though it requires more integration work than the native Marketing Cloud Connect for Sales Cloud. EPC Group builds integrations using MuleSoft, Azure Logic Apps, or custom REST API connections to synchronize contacts, leads, campaign data, and engagement metrics between Marketing Cloud and Dynamics 365.

For Microsoft-committed organizations, we also evaluate Dynamics 365 Customer Insights (Marketing) as a potential alternative.

What scripting languages does Marketing Cloud use?

Marketing Cloud uses three languages: AMPscript for inline email and landing page personalization, Server-Side JavaScript (SSJS) for complex logic in CloudPages and Script Activities, and SQL for data queries in Automation Studio.

REST and SOAP APIs handle external integrations from any programming language. EPC Group developers are proficient in all three and commonly build external integrations in Node.js, Python, and C#/.NET.

How long does a Marketing Cloud implementation take?

A typical implementation takes 2–4 months for the foundation (data model, integration, basic campaigns) and 4–8 months for full maturity (advanced journeys, multi-channel, AI optimization).

Complex multi-Business Unit deployments with custom integrations can take 6–12 months. EPC Group's agile approach deploys core email capabilities within the first 4–6 weeks, then adds capabilities in iterative phases.