Tableau teams have spent years building trusted, governed data sources around specific business needs. But the moment a new question reaches beyond the boundaries of one of those sources, analysts often find themselves back at the beginning: requesting changes, recreating logic, or building another version of something that already exists.
Tableau’s Composable Data Sources create new options for data modelers. Beginning with Tableau 2026.2, the feature allows published data sources to become reusable building blocks that can be combined and extended for new analytical needs. Instead of rebuilding the model, teams can build on what they already trust.
What Are Tableau Composable Data Sources?
Composable Data Sources expand what teams can do with existing published data sources through two related capabilities: composing and extending.
Composing allows authors to relate two or more published data sources within the same data model, then use that model in a workbook or publish it as a new reusable source. The original sources remain separate and unchanged.
Extending allows authors to customize an existing published source with changes such as calculations, renamed or hidden fields, groups, bins, and other supported overrides. Those changes can also be published as a new source without altering the original.
Together, these capabilities allow teams to build purpose-specific data models from trusted assets rather than continually recreating them.
From Fixed Endpoints to Reusable Building Blocks
The real value of composability is that published data sources no longer have to represent the end of the modeling process. Teams can keep trusted sources focused on their intended purpose, then reuse them as building blocks when a new analytical question requires additional context.
A sales model, for example, can remain focused on sales while other published sources provide targets, customer engagement, fulfillment activity, financial data, or a shared calendar. Authors can relate those sources at the appropriate grain and create a model for the question at hand without rebuilding the underlying logic.
This approach preserves the investment already made in governed data while giving teams more flexibility to respond to new analytical needs. Upstream sources remain intact, and changes made to an extended source do not alter the originals.
How Composable Data Sources Change the Tableau Workflow
Composable Data Sources introduce a more flexible way to approach new analytical requirements. Instead of beginning with the question, “What do I need to rebuild or add to this source?” authors can first ask, “What trusted data already exists that I can use?”
Tim Ngwena’s walkthrough of composable data sources demonstrates this shift well. Starting with an established sales model, he composes in additional published sources for checkout activity, shared dates, sales targets, and row-level security. Each source contributes something specific to the analysis while the original sales model remains intact.
The result is a more modular workflow: start with trusted sources, compose the pieces needed for the current question, extend the model where necessary, and publish the result for broader reuse when it has value beyond a single workbook.
Why Composable Data Sources Matter
For business and data leaders, the most significant benefit may be a lower cost of change. New questions no longer have to trigger a new modeling effort simply because the required data crosses the boundaries of an existing source.
Composable Data Sources give teams more freedom to respond to changing business needs while preserving the governed assets already in place. Analysts can bring trusted sources together for a new purpose, data teams can avoid unnecessary duplication, and existing consumers can continue using the original models without disruption.
That creates a better balance between governance and agility. Organizations can maintain focused, trusted sources while giving analysts a practical way to extend beyond them when the business question demands it.
Lower Cost of Change → Faster Answers → Governance Without Rigidity
Governance Becomes an Enabler
Composable Data Sources do not reduce the need for governance. They make good governance more valuable.
When trusted data sources can be reused across more models and use cases, clear ownership, naming, certification, and documentation become essential. Teams need to know which sources are authoritative, what they contain, who maintains them, and how downstream models depend on them.
With those standards in place, governance becomes an enabler rather than a constraint. Analysts gain more flexibility to compose the data they need, while organizations preserve consistency, lineage, and trust across the environment.
Good governance does not compete with composability. It is what makes composability scalable.
What to Know Before You Compose
Composable Data Sources are new in Tableau 2026.2, and the first release has some important boundaries to understand before putting them into production.
- Composition begins with published data sources. Local files must be published before they can participate in a composed model.
- Permissions carry through the composition. Users need access to every upstream published source. There is no partial access to a composed data source.
- Live and extract workflows differ. In 2026.2, only live composed data sources can be published as standalone data source assets. Extracted compositions can be published within a workbook, with the extract created in Tableau Desktop.
- Connector support is not universal. Some connectors are not currently compatible, including Salesforce, so confirm support for the sources you intend to use.
- Performance should be tested. Tableau currently uses cross-database joins for composed sources, even when underlying sources use the same connection and credentials.
Tableau also applies safeguards around the number and depth of upstream dependencies. Most implementations are unlikely to approach those limits, but they reinforce a useful principle: compose intentionally around a business need rather than creating complexity simply because the architecture allows it.
Because these capabilities are evolving, review Tableau’s current documentation before moving a composed model into production.
A Practical Way to Get Started
The best place to begin is with a focused use case: a valuable business question that already requires analysts to cross the boundaries of two or three trusted published data sources.
Start with a real business question, not a composability experiment.
Start by confirming the sources are well understood, governed, and appropriate to relate. Then build the composed model, validate the relationships, permissions, performance, and business logic, and publish it with clear ownership and documentation if the result has value beyond a single workbook.
The goal of the pilot is not simply to prove that composability works. It is to determine whether the approach reduces rework, improves consistency, and helps teams reach trusted answers faster.
Compose with Purpose
Composable Data Sources give Tableau teams a practical way to move faster without giving up trust. The opportunity is not to compose everything, but to identify where trusted, existing data sources can be reused to answer new questions with less rebuilding, less duplication, and greater flexibility.
If your team is exploring where composability fits in your Tableau environment, XeoMatrix can help identify strong use cases, establish governance standards, and design a data source strategy that can grow with your needs. Let’s talk about what you could compose next.
