Discover Best Practices for Tableau Next Workspaces

Learning Objectives

After completing this unit, you’ll be able to:

• Identify best practices for structuring workspaces.
  
• List different types of collaboration models for data analysis.
  

Optimize Your Workspace Strategy

Before you dive into the technical build, it’s important to understand why you’re creating a workspace and how it will serve your team. In Tableau Next, a workspace is more than just a folder; it’s a modular environment designed to streamline how you curate data, build visualizations, and leverage AI capabilities. By intentionally structuring your workspaces, you create a foundation for seamless collaboration and governed data discovery.

Apply Best Practices for Building Workspaces

As you start building workspaces in Tableau Next, these questions can help you think through how to use this flexible, modular environment to your advantage.

• Purpose: What’s the overall objective of your work?
  
• Collaboration: If you plan on having multiple contributors, how will the work be partitioned?
  
• Governance: Are there any aspects of the analytic workflow that would make sense to be targeted for reuse and more governance?
  
• AI optimization: How can you set up your workspace to get the most out of the AI-powered features in Tableau Next?
  

Let’s take a closer look at each of these factors to help you determine how to organize your team’s workspaces.

Define the Purpose

It’s best to design your workspace based on your team’s specific goals.

Maybe that means creating a new workspace for each discrete project or analytical question. This helps keep your analytical assets organized and focused. Even if some assets are similar, a new workspace might be a good idea if the analytical question is different.

You can also differentiate between quick investigations and building structured assets.

• For ad-hoc investigations (such as finding a root cause, gathering insights for a business decision, or testing a dataset), you can create a new, temporary workspace to capture all related assets and experiments, then use another workspace to build specific assets when the investigation work is complete.
  
• For building specific assets (like a dashboard or a semantic model with clear specifications), the workspace should contain all directly related building blocks, including attempts, data validations, or failed experiments.
  

Identify Your Team’s Collaboration Style

When you’re building workspaces, keep collaboration in mind and consider ways to best support your team’s working style. Every team works together differently (and that’s OK!). Here are some recommendations for a few of the most common team dynamics.

• “Wild West” or “Anything goes”: For teams who collaborate heavily, all contributors can work on any asset within a single workspace.
  
• Vertical partitioning: Each contributor takes end-to-end responsibility for a specific part of the work (for example, one team member manages the visualization while another curates the dashboard).
  
• Horizontal partitioning: Divide work by analytical function or “swimlanes.” For instance, one person can curate the data in their own workspace, and then share that data asset (such as a semantic model) with others. Other team members can then reference this shared asset in their separate workspaces to build their analytical content (like dashboards or visualizations). This approach allows others to benefit from ongoing data improvements without managing the curation themselves.
  
• Combined partitioning: For larger teams, a hybrid approach can be used, where a central team creates and shares a core semantic model, and other teams reference that model to build their own dashboards. This approach allows for a complete overview while maintaining separate work areas.
  

Ensure Data Governance

Data governance is essential for ensuring that analytical assets are reusable and well-managed. One key way to achieve this is by creating workspaces with specific ownership. This allows different teams or individuals to be responsible for managing particular assets, which promotes a clear structure, and makes it easier to track and govern the entire analytic workflow. By focusing on governance at this level, you can ensure consistency and reliability across all shared assets.

Sometimes, only a small number of people should have access to certain analytical assets. For example, if your team handles sensitive financial models, you can create a dedicated workspace with controlled access for your critical content.

You can then share this workspace with view-only permissions for other analysts. By controlling who can modify the workspace, you ensure that only your team can make changes, guaranteeing that everyone in the organization uses trusted and accurate data.

Remember that access to data model objects (DMOs) and data lake objects (DLOs) is controlled by policies in Data 360, not directly by sharing a workspace. To give users access to a semantic model, add them to the data space the model belongs to before sharing the workspace with them.

Build with AI Optimization in Mind

Tableau Next is an AI-powered analytics platform, and how you set up your semantic models is the key to getting the most out of it. By following these best practices, you ensure that AI features like Concierge and Inspector can work their magic.

• Structure your data for clarity: Build a semantic layer that includes all your business logic and definitions. It’s the foundation the AI relies on.
  
• Use business-friendly names: Give your fields and calculations clear, simple names like Product Name instead of technical ones like Prd_id. Add short descriptions to explain their purpose.
  
• Define your metrics: Make sure all your key performance indicators (KPIs) are clearly defined as metrics. The more metrics you have, the more trusted answers Concierge can give you.
  
• Avoid ambiguity: Help the AI understand your data by clearly naming similar fields and metrics. For example, search for generic terms like “customer” and rename or remove any fields that could cause confusion.
  
• Validate relationships: Double-check that all the relationships between your data are logical. This helps the AI connect the dots correctly.
  
• Enable AI readiness: Once a semantic model follows these best practices, turn on the Agentforce for Analytics Ready flag. This tells Concierge it’s a trusted source it can use to answer questions.
  
• Be mindful of costs: To keep costs low, include only the data you need for a specific use case. You can reduce data rows and manage refresh schedules to make sure you’re not over-querying.
  

By thoughtfully structuring your Tableau Next workspace, you can help your team work more effectively and gain deeper insights.