Skip to main content
tableau semantic layerTableau SemanticsTableau NextTableau Pulsepublished data sourcesData 360semantic layersemantic model

What Is the Tableau Semantic Layer? Tableau Semantics

The Tableau semantic layer is the set of Tableau features that define what data means before anyone builds a chart: the data model inside published data sources, the metric definitions behind Tableau Pulse, and Tableau Semantics, the semantic model engine that powers Tableau Next on the Salesforce platform. Together they hold the relationships, calculations, and business names that keep a number consistent across workbooks, dashboards, and AI agents.

The term is confusing because Tableau has had semantic features for years without calling them a semantic layer, and because the newest one, Tableau Semantics, lives in a different product. Which layer a team means depends on whether it works in Tableau Cloud, Server, and Desktop or in Tableau Next. This article explains all three and how they connect.

TL;DR

Classic Tableau keeps meaning in published data sources: a two-layer data model (relationships in the logical layer, joins and unions in the physical layer), calculated fields, and names shared across workbooks. Tableau Pulse adds a metrics layer of metric definitions built on those data sources. Tableau Semantics is the semantic layer of Tableau Next, built on Data 360 (renamed from Data Cloud in October 2025); its semantic models combine a data model with business definitions and feed metrics, dashboards, and Agentforce agents, and Tableau Desktop 2025.2 and later can connect to them. Tableau can also read other semantic layers, such as the dbt Semantic Layer and Snowflake semantic views. Ownership and lineage across platforms come from a context layer.

What Is the Tableau Semantic Layer?

A semantic layer translates tables and columns into business terms and calculations that people and tools can reuse. In Tableau that job is split across three features:

  • The data model in published data sources, available in Tableau Cloud, Tableau Server, and Tableau Desktop.
  • Tableau Pulse metric definitions, a metrics layer on top of published data sources.
  • Tableau Semantics, the semantic model technology of Tableau Next and Data 360.

All three aim at the same outcome, one definition of a measure that every view uses, but they differ in scope, in who maintains them, and in which product reads them.

The Tableau Data Model and Published Data Sources

Every Tableau data source has a data model that tells Tableau how to query the connected tables. Since version 2020.2 the model has two layers:

  • The logical layer, where tables are combined with relationships. You do not choose a join type; Tableau picks the join at query time based on the fields in the view, which avoids many duplication problems in star and snowflake schemas.
  • The physical layer, where each logical table can be built from one or more physical tables through joins and unions.

When a data source with its relationships, calculated fields, default aggregations, and friendly field names is published to Tableau Cloud or Server, every workbook connected to it inherits the same logic. For many organizations this has long been their Tableau semantic layer in practice. Its limits are that each published data source is its own island, definitions are edited in Tableau's UI rather than in code, and the logic is only reused by tools that query through Tableau.

Tableau Pulse Metric Definitions

Tableau Pulse monitors metrics in the flow of work, with digests, insights, and alerts. Its metrics layer is described by Tableau as a single source of truth for your data. A metric definition holds the core metadata a metric is based on: the measure, the aggregation, the time dimension, and the filters people can adjust. Individual metrics are then created from a definition by changing filters and time options for different groups of followers.

Metric definitions are built on published data sources, and the permissions on a data source control who can create definitions from it. That makes Pulse a metrics layer that sits on top of the data model described above, focused on tracking a number over time rather than on general-purpose analysis. For the general concept, see what a metrics layer is.

What Is Tableau Semantics?

Tableau Semantics is the technology behind semantic models in Tableau Next, the API-first analytics platform Salesforce builds on Data 360. Data 360 is the new name for Salesforce Data Cloud as of October 14, 2025, so older material uses both names. In Tableau Next, the semantic models form the semantic layer, and Tableau Semantics also handles the queries needed for analytics.

A Tableau Next workspace moves through layers:

  • Data: always a Data 360 object, such as a data model object, or a CSV stored in Data 360.
  • Semantic models: the data model (which objects are used and how they relate) plus business definitions such as field names, aggregations, and calculations.
  • Metrics: built on semantic models to track measures over time.
  • Vizzes and dashboards built on top of those.

Semantic models are created in the Semantic Model Builder, available in Tableau Next and in Data 360. Tableau Next can also create a semantic model from a published data source in Tableau Cloud, which links the classic and the new semantic layer. In the other direction, the Tableau Semantics connector, released with Tableau Desktop 2025.2 and available in Desktop and Tableau Cloud, lets classic Tableau use a semantic model as a data source. The connector retrieves the semantic metadata as well as the data, works only live (no extracts), and gives the model's definitions precedence over local changes.

The Tableau Semantic Layer THE TABLEAU SEMANTIC LAYER - TWO PRODUCTS, ONE GOAL TABLEAU CLOUD, SERVER, DESKTOP Published data sources Relationships, joins, unions Calculated fields, field names Tableau Pulse metric definitions (measure, aggregation, time, filters) Meaning shared across workbooks TABLEAU NEXT ON DATA 360 Tableau Semantics Semantic models: data model plus business definitions Metrics, vizzes, dashboards Agentforce agents Built in Semantic Model Builder Model from a published data source Tableau Semantics connector (2025.2+) THE CONTEXT LAYER AROUND IT One business glossary with approved definitions and named owners for Tableau and every other tool Lineage from source systems and warehouses to the Tableau workbooks that show the number Tableau defines how a measure is calculated; the context layer says what it means and who owns it
Click to enlarge

Tableau Semantics and AI Agents

Salesforce positions Tableau Semantics as the foundation for agentic analytics. In Tableau Next, Agentforce tools such as Concierge (questions in natural language), Inspector, and Data Pro work on top of semantic models, so an agent answers from the model's definitions instead of guessing joins and calculations from raw objects.

For AI tools outside Salesforce, Tableau publishes an official Tableau MCP server. It can run as a hosted service for Tableau Cloud, where each user signs in with their own Tableau identity through OAuth and their existing permissions apply, or self-hosted against Tableau Cloud or Tableau Server. Through it, an MCP client can query published data sources and find content, which means the semantic work already done in those data sources becomes available to assistants such as Claude.

The same caution applies as with any semantic layer: an agent reading a Tableau model gets a consistent answer, which is only a correct answer if the definition in the model is the one the business agreed on.

Tableau With Other Semantic Layers

Many teams keep metric definitions upstream of Tableau and let Tableau read them. Documented options as of November 2026:

  • dbt Semantic Layer: Tableau is one of the native integrations listed by dbt, so Tableau users can query MetricFlow metrics. See the dbt Semantic Layer.
  • Snowflake semantic views: Snowflake can export a semantic view as a Tableau Data Source (TDS) file, in public preview since March 2026, used through a live connection. Snowflake also lists Tableau among the BI tools adopting its semantic definitions. See Snowflake semantic views.
  • Databricks metric views: Tableau can query them with Custom SQL and the MEASURE() function, or with BI compatibility mode, and Databricks has announced a beta import of Tableau models into metric views. See the Databricks semantic layer.
  • Looker: Looker offers a BI connector for Tableau Desktop, so LookML models can be queried from Tableau. See the Looker semantic layer.

The pattern is the same everywhere: the closer the definition sits to the data, the more tools can share it, and the less each BI tool has to own.

Tableau Semantic Layer vs Context Layer

LayerWhat it definesWho reads itWhere it stops
Published data sourceRelationships, joins, calculated fields, field namesTableau workbooks, Pulse, Tableau MCPOne data source at a time, inside Tableau
Pulse metric definitionsMeasure, aggregation, time dimension, filtersPulse followers and digestsMetrics tracked in Pulse
Tableau SemanticsSemantic models over Data 360 objectsTableau Next, Agentforce, Tableau Desktop via connectorData brought into Data 360
Context layerGlossary terms, approved definitions, owners, lineage, quality signalsPeople and AI agents across all toolsSpans warehouses, BI tools, and source systems

Tableau's semantic features decide how a measure is calculated inside Tableau. A context layer records what the business means by that measure, who owns the definition, and where the data comes from, for Tableau and for every other tool. More in context layer vs semantic layer and ontology vs semantic layer.

How Dawiso Works With Tableau

Dawiso connects to Tableau Cloud and Tableau Server through the Tableau connector, which uses the Tableau REST API and the Metadata (GraphQL) API to retrieve site assets and their relations. The Metadata API is on by default in Tableau Cloud and has to be enabled on Tableau Server, and Dawiso sees only the Tableau assets visible to the integration user. Tableau sits next to more than 30 other sources Dawiso reads, including Snowflake, Databricks, Microsoft Fabric, and Power BI.

That puts Tableau content into the same catalog as the warehouses and pipelines behind it. The business glossary holds the approved definition and owner of a metric that a Tableau calculated field or Pulse definition implements, and lineage connects the workbook to its sources. AI tools that support MCP can query Dawiso through the Dawiso MCP Server and get the definition, owner, and lineage along with the number. For other semantic layer options, see the best semantic layer tools.

Conclusion

Tableau's semantic layer is not one feature. Published data sources and Pulse metric definitions carry meaning in classic Tableau, and Tableau Semantics carries it in Tableau Next on Data 360, with a connector and model creation linking the two. Tableau can also read definitions kept upstream in dbt, Snowflake, Databricks, or Looker. Whichever layer calculates the number, the business definition behind it, its owner, and its lineage across platforms belong in a context layer that every tool and agent can reach.

Sources

See it in action

Business Glossary

Clear context is essential to ensure everyone interprets terms consistently and accurately.

Table of contents

What Is the Tableau Semantic Layer?Data Model and Published Data SourcesTableau Pulse Metric DefinitionsWhat Is Tableau Semantics?Tableau Semantics and AI AgentsTableau With Other Semantic LayersTableau Semantic Layer vs Context LayerHow Dawiso Works With TableauSources

Related Terms

What Is the Looker Semantic Layer? LookML Explained

LookML was built for Looker dashboards. Now the Looker semantic layer also feeds Gemini, AI agents through MCP, and BI tools such as Tableau and Power BI.

Read more →

What Are Snowflake Semantic Views & Cortex Analyst?

Snowflake semantic views are native objects defining metrics, relationships, and synonyms so Cortex Analyst turns plain-English questions into accurate, governed SQL.

Read more →

What Is a Power BI Semantic Model (Formerly Dataset)?

A Power BI semantic model, formerly a dataset, holds the tables, relationships, DAX measures and security rules that reports, Copilot and AI agents reuse.

Read more →

What Is the dbt Semantic Layer? MetricFlow Explained

The dbt Semantic Layer defines metrics once in YAML next to your dbt models, and MetricFlow turns them into SQL for BI tools, spreadsheets, and AI agents.

Read more →

What Is the Databricks Semantic Layer? Metric Views

Two dashboards, two revenue numbers? The Databricks semantic layer fixes it with Unity Catalog metric views that Genie, SQL, and BI tools all reuse.

Read more →

Semantic Layer vs Traditional Data Marts

Semantic layer vs data marts - how a central semantic layer defines each metric once for every connected tool, versus per-team marts that define their own.

Read more →

A cookie a day keeps bad UX away.

We use cookies to personalize content, ads and to analyze our traffic. We also share information about your use of our site with our advertising and analytics partners who may combine it with other information that you've provided to them or that they've collected from your use of their services. By clicking "Accept All", you allow us to use cookies for analytics and ads via Google Tag Manager. You can also customize cookies.

Customize Consent Preferences

We use cookies to personalize content, ads and to analyze our traffic. We also share information about your use of our site with our advertising and analytics partners. Privacy Policy

Necessary cookies allow core website functionality such as user login and account management. The website cannot be used properly without strictly necessary cookies.

Functionality cookies are used to remember visitor information on the website, eg. language, timezone, enhanced content.

Analytics cookies are used to see how visitors use the website, eg. analytics cookies. Those cookies cannot be used to directly identify a certain visitor.

We use Microsoft Clarity to see how you use our website (including heatmaps and session replays) so we can improve it. By using our site, you agree that we and Microsoft can collect and use this data. See our Privacy Policy for details.

Advertisement cookies are used to identify visitors between different websites, eg. content partners, banner networks. Those cookies may be used by companies to build a profile of visitor interests or show relevant ads on other websites.