Unified Marketing Intelligence Semantic Data Model

How SDM Queries Work

To query the SDM, choose the dimensions for grouping or filtering, the measurements for analysis, and the date range.

• Measurements define what you want to count, sum, average, or calculate, such as impressions, clicks, spend, opportunities, or revenue.
• Dimensions define how you want to group or filter results, such as by Campaign, channel, Market Segment, site, or opportunity stage.
• Date range defines the reporting period, such as this quarter, last month, or a custom date range.

For example, to answer “How did each campaign perform this quarter?”, use campaign performance measurements, group the results by Campaign, and set the date range to this quarter.

Example query goals include:

• Show impressions, clicks, spend, and CTR by Campaign for this quarter.
• Compare paid media clicks and owned-channel clicks by Campaign.
• Show opportunity pipeline value and campaign spend by Campaign.
• Analyze website visits and landing page views by Campaign.
• Review delivery quality by ad, media buy, or site.

Modules at a Glance

The SDM includes three analytics modules and one unified modeling layer.

Cross-Module Querying Rules

Measurements from different modules can be combined only when the modules share a dimension.

Campaign is the universal shared dimension. Use Campaign to report across MI, MP, and B2B. Market Segment is shared by MI and MP only. You can use Market Segment to compare paid media and owned-channel engagement, but not B2B pipeline.

Most other dimensions are module-specific. For example, Media Buy, Ad, Journey, Opportunity Stage, and Website Page are useful within their own modules, but they don’t join all modules. Subscription and consent measurements are isolated. Don’t combine them with paid media, engagement, or pipeline measurements unless a supported shared dimension is available.

Architecture

The SDM is generated from governed model definitions and deployed to Salesforce Data Cloud in three layers: Base, System, and Custom. Deploy the layers in that order.

Base layer

Contains the foundational model, including core entities, logical views, raw measurements, relationships, and channel-specific dimensions.

System layer

Extends the Base layer with shared reporting logic, unified dimensions, compound logical views, and cross-channel measurements.

Custom layer

Supports customer-specific extensions that build on the System layer. Add customizations here instead of changing generated Base or System definitions.

Marketing Intelligence

Marketing Intelligence models paid media performance across integrated ad platforms. It includes delivery, conversion, verification, and web analytics data. Use MI to analyze paid media performance by campaign, media buy, ad, site, device, market segment, and other paid media dimensions.

Each MI engagement type connects to Campaign through its own Media Buy Path logical view. Delivery also connects to Marketing Channel, Media Buy Package, Product, and Site.

MI Relationships and Join Paths

Marketing Intelligence data is organized around engagement tables and dimension entities. Engagement tables store measured activity, such as delivery, conversion, verification, and website activity. Dimension entities provide reporting context, such as Ad, Device, Market Segment, Media Buy, Campaign, Product, and Site.

Media Buy connects paid media activity to the campaign hierarchy. Each MI engagement type has a Media Buy key and connects to Campaign through a dedicated Media Buy Path logical view.

Key MI Measurements

Marketing Intelligence includes raw measurements and calculated measurements for paid media delivery, conversions, verification, and web analytics.

Paid Media Delivery Measurements

Conversion and Revenue Measurements

Verification and Web Analytics Measurements

Calculated Measurements

Calculated measurements are derived from raw measurements after filtering and grouping so that rates and ratios reflect the selected reporting context.

Note

For ratio measurements, the SDM returns no value when the denominator is zero. For example, CTR is calculated only when impressions are greater than zero.

Key MI Dimensions

Conversion Model

The Conversion model organizes conversion and revenue data from integrated platforms. It supports post-click and post-impression attribution when those values are available from the source platform.

Delivery Model

The Delivery model organizes advertising delivery and engagement performance across paid and organic channels. It covers impressions, clicks, social engagement, video, search, rich media, and cost measurements.

Delivery Verification Model

The Delivery Verification model organizes measurements for ad viewability, verification integrity, and delivery quality. Use it to analyze ad delivery quality, identify verification issues, and compare performance across measurement standards.

Web Analytics Event Model

The Web Analytics Event model tracks user interactions on a website, such as form submissions, button clicks, downloads, and other configured events.

Web Analytics Page Model

The Web Analytics Page model provides a session-level view of website traffic, retention, and on-site financial performance. Use it for visit quality, bounce, engagement, and transaction reporting.

Marketing Performance

Marketing Performance represents engagement across owned channels, including email, SMS, RCS, WhatsApp, push notifications, in-app messages, websites, and landing pages. Use MP to analyze how customers interact with owned marketing messages and digital experiences.

The four engagement tables are the heart of MP. Each engagement table connects owned-channel activity to message, individual, website publication, and campaign context.

Marketing B2B

Marketing B2B adds sales pipeline and opportunity analytics to the SDM. Use B2B to analyze opportunities, revenue, accounts, contacts, and sales activity.

Opportunity is the hub of the B2B relationship graph, linking to Account, Account Calendar Event, Account Contact, and User. Opportunity Influence connects Opportunity to Campaign. Campaign is B2B’s bridge to the rest of the model. Because Campaign is the only join path from B2B to MI and MP, pipeline measurements can be grouped by Campaign alongside paid and owned-channel measurements. B2B doesn’t include a Market Segment join, so pipeline measurements can’t be grouped or filtered by Market Segment.

Opportunity Measurements

Campaign Investment Measurements

Conversion Timing Measurements

Sales Activity and Account Measurements

Note

B2B measurements are usually grouped by Campaign through Opportunity Influence, or by opportunity and account dimensions within the B2B module. Use Campaign when combining B2B pipeline measurements with MI or MP marketing engagement measurements.

Unified Modeling

The Unified model combines MI, MP, and B2B into one deployable model. It supports cross-channel reporting by using shared dimensions and unified calculated measurements.

Campaign is the primary shared dimension across all three modules. Use Campaign to compare paid media, owned-channel engagement, and B2B pipeline measurements in the same query. Market Segment is shared by MI and MP only. B2B doesn’t join to Market Segment, so pipeline measurements can’t be grouped or filtered by Market Segment.

Shared entities such as Campaign, Market Segment, Account, and Opportunity combine dimensions from participating modules. When the same field exists in multiple modules, the unified model applies a consistent conflict-resolution rule so the reporting model has one usable field.

Unified Calculated Measurements

Unified Dimensions

Different engagement types can store the same concept under different field names. Unified dimensions create virtual columns across engagement types by using union logical views.

Cross-Module Query Examples

Campaign performance across marketing and pipeline

Use Campaign when comparing all three modules.

Measurements

Paid media impressions, paid media spend, owned-channel clicks, opportunity pipeline value, and closed won opportunities

Dimension

Campaign

Date range

This quarter

Engagement by market segment

Use Market Segment when comparing MI and MP only.

Measurements

Paid media impressions, paid media clicks, email clicks, and SMS clicks

Dimension

Market Segment

Date range

Last month

Don’t use Market Segment for B2B pipeline reporting unless the model exposes a supported B2B relationship for that workspace.

Implementation Rules for Reporting

• Use SUM for raw counts and totals such as impressions, clicks, spend, conversions, sends, opens, and revenue.
• Use AVG when the source field already stores a precomputed average, such as frequency, search impression share, or average display position.
• Use UserAgg for ratios and rates that must be calculated after filtering and grouping, such as CTR, CPC, ROAS, Bounce Rate, and Average Time on Site. This method prevents errors caused by averaging row-level percentages.
• Use count distinct for unique entity counts, such as unique campaigns or unique market segments.

Directionality Reference

Directionality indicates whether higher or lower values usually represent better performance.

Spend can be interpreted differently depending on the analysis. In budget and investment reporting, higher spend can represent more investment. In efficiency reporting, lower spend for the same outcome is usually better.

Null Values

Measurement fields use semantic null handling so missing measurement values are treated as zero in reporting calculations. This behavior helps calculations such as CTR, ROAS, and Bounce Rate avoid errors when a source platform doesn’t provide data for a segment. It also helps reports display zero instead of blank values where appropriate.

Glossary

Campaign

A shared dimension used to group and compare marketing and pipeline activity across MI, MP, and B2B.

Calculated Dimension

A text or classification field derived from model logic.

Calculated Measurement

A numeric measurement, such as a rate or ratio, derived from one or more base measurements.

Dimension

A descriptive field used to group or filter measurements.

Engagement

A fact record for a measured interaction, such as an ad impression, conversion, email click, message send, or website visit.

Logical View

A virtual view that combines or relates underlying entities for reporting.

Market Segment

A shared MI and MP dimension used for audience or segment-level analysis.

Measurement

A value that can be counted, summed, averaged, or calculated.

Metric

A reporting-ready measurement that supports aggregation and analysis over time.

Semantic Data Model

A governed reporting model that defines entities, dimensions, relationships, and measurements for analytics.

Unified Dimension

A dimension that standardizes the same reporting concept across multiple engagement types or modules.