How to Create a Data Source in Tableau Server

Creating a centralized, trustworthy data source is one of the most effective ways to level up your team’s analytics game. This article will walk you through exactly how to create and publish a reusable data source in Tableau Server, transforming raw data into a reliable asset for your entire organization.

What is a Published Data Source (and Why Should You Care)?

In Tableau, you typically start by connecting Tableau Desktop to some data - maybe an Excel file, a Google Sheet, or a SQL database. A published data source is what happens when you take that connection, clean it up, and then publish it to your Tableau Server for others to use.

Instead of ten different analysts all connecting to the same messy database and creating their own slightly different calculations for "Revenue," you can do it once. You create the definitive "Revenue" calculation, hide unnecessary columns, and publish it as the official "Company Sales Data" source. Now, anyone building a new sales report can just connect to that curated source, saving time and ensuring everyone is working from the same numbers.

The benefits are huge:

• A Single Source of Truth: This is the big one. It guarantees consistency. When the sales and marketing teams both use the same published data source, their reports will always align. No more meetings spent arguing about whose numbers are correct.
• Improved Performance: You can create and schedule a data extract on the server. This means Tableau pre-packages the data into a high-performance snapshot, making dashboards load much faster because they aren't querying the original database live every time someone opens them.
• Enhanced Security: You can set permissions on the data source itself. This allows you to control who can view or connect to sensitive data. You can also embed database credentials, so users don't need direct access to the underlying database, which keeps your data more secure.
• Increased Efficiency: Analysts don’t have to waste time reinventing the wheel. They can skip the data connection and cleaning steps and get straight to building insightful visualizations. New team members can become productive immediately by using pre-built, sanctioned data sources.

Prerequisites: What You’ll Need Before You Start

Before you get started, make sure you have a few things in place. This will save you from hitting roadblocks later.

• Tableau Desktop: Publishing is done from the Tableau Desktop application, not the web interface. Make sure you have it installed.
• A Tableau Server Creator License: Only users with a "Creator" role have the permissions to publish new content, including data sources, to Tableau Server. A "Viewer" can't publish, and an "Explorer" has limitations.
• Server/Site Access: You'll need the URL for your Tableau Server and a username/password with publishing rights to a specific project folder. If you're not sure where to publish, check with your Tableau administrator.
• Database Credentials: You’ll need the necessary credentials (server name, username, password) to connect to the raw data a single time from your desktop to set everything up.

Step-by-Step: Creating and Publishing Your First Data Source

Let's walk through the process using a common scenario: building a master sales data source by combining order information from a transactions table with customer details from a customer table.

Step 1: Connect to Your Data in Tableau Desktop

The first step is always connecting Tableau Desktop to your raw data.

1. Open Tableau Desktop.
2. In the Connect pane on the left, choose the appropriate connector. If you are connecting to a SQL database, you would select connectors like Microsoft SQL Server or PostgreSQL. For this example, let's say our data is in a set of related tables in a single Excel file. We'll click on Microsoft Excel.
3. Navigate to your file and click Open.

Once connected, you'll be automatically taken to the Data Source page, which is your workshop for preparing the data.

Step 2: Prepare and Clean Your Data

This is where the magic happens. On the Data Source page, you can shape the raw data into a clean, easy-to-use model before publishing.

In the top left pane, you'll see the tables (or sheets, for an Excel file) from your data connection. Let's start building:

Create Relationships or Joins

Drag the tables you need into the canvas area. For our example, let's drag out the "Orders" table first. Then, drag out the "Customers" table. Tableau will often automatically detect the common field (like "Customer ID") and create a relationship, which looks like a thin line or "noodle" between the tables. This allows you to analyze data from both tables together.

Curate Your Fields

Your tables might have dozens of columns that aren’t useful for general analysis. Hiding them makes the data source much cleaner for end-users.

• Below the canvas, you’ll see a preview of your data grid. You can click the small downward arrow on a column header and select Hide to remove it from the view.
• You can also rename fields to be more user-friendly. For example, you can right-click the "Cust_Name" column and rename it to "Customer Name."

Create Calculations

If there's a key metric everyone on your team uses, define it here once. Let's create a "Sales per Order" calculation.

1. Navigate to any worksheet (by clicking the "Sheet 1" tab at the bottom).
2. In the Data pane on the left, click the dropdown arrow at the top and select Create Calculated Field.
3. Name the field "Sales per Order".
4. Enter the formula:

SUM([Sales]) / COUNTD([Order ID])

1. Click OK. Now this calculation is part of your data source and will be available to anyone who uses it.

After your calculations and cleaning are done, navigate back to the Data Source tab.

Step 3: Choose Between a Live Connection and an Extract

At the top right of the Data Source page, you have a critical choice: Live vs. Extract.

• Live: A live connection queries your database directly. Any changes in the source data will be reflected in your dashboard in real-time. This is great for operational data that needs to be constantly up-to-date, but dashboard performance depends entirely on your database's speed.
• Extract: An extract takes a snapshot of your data and stores it in Tableau’s optimized, file-based data engine (.hyper format). This results in much faster performance for large datasets. You can then schedule the extract to be refreshed on a recurring basis (e.g., daily, hourly) directly on Tableau Server.

For most analytics use cases where real-time data isn't a strict requirement, an Extract is the best practice for performance. Select the Extract radio button. Tableau will prompt you to save the extract file when you navigate to a sheet, which is a necessary intermediate step before publishing to the server.

Step 4: Publish the Data Source to Tableau Server

With your data prepared and your connection type chosen, you're ready to publish.

1. From the top menu, go to Server > Publish Data Source.
2. If you're not already signed in, Tableau will prompt you to do so. Enter your Tableau Server URL and your credentials.
3. The Publish Data Source dialog box will appear. This is your final checkpoint.

Here’s a breakdown of the key settings:

• Project: Choose the folder on Tableau Server where this data source will live. Organizing data sources into projects (e.g., "Marketing," "Sales," "Finance") is essential for keeping your server tidy.
• Name: Give it a clear, descriptive name like "Sales & Customer Master Data." This is what users will see when they look for a data source on the server.
• Description: Always add a good description. Explain what the data contains, what key calculated fields mean, and who the owner is. This is invaluable for other users.
• Add Tags: Use keywords like "Sales," "Orders," or "E-commerce" to make your data source easier to find via search.
• Authentication: This tells Tableau Server how to handle credentials for the underlying database.

When everything is configured, click the blue Publish button. It may take a few moments for Tableau to create the extract (if selected) and upload it to the server.

Best Practices for a Healthy Tableau Server Environment

Just publishing a data source isn't enough. Managing it effectively is what truly unlocks its value.

• Develop a Naming Convention: Don't just call it "Data." Be specific. A good practice is something like [Department][Content][Type], e.g., Sales_OrdersCurrentYear_Extract.
• Set Refresh Schedules Wisely: If you published an extract, go to Tableau Server, navigate to your data source, and set up a refresh schedule. Be mindful of server resources - does this really need to be refreshed every 30 minutes, or is once a day sufficient?
• Manage Permissions: Use Tableau Server's permissions to control who can do what. For example, you might let everyone in a certain group connect to the data source, but only allow a few specific people the ability to edit or overwrite it.
• Update Descriptions: If you add a new crucial metric or make a structural change, update the description on the server to reflect it. Treat the data source like a living product for your internal team.

Connecting to Your New Published Data Source

Now for the payoff! Once published, any user with permission can connect to this source from their own Tableau Desktop instance to create new reports and dashboards.

1. In Tableau Desktop, under the Connect pane, select Tableau Server.
2. Navigate to the project and select the published data source you just created.
3. Click Connect. That's it!

You'll see all your cleaned-up fields, pre-built calculations, and proper data types, ready to go. You just provided a governed, high-performance data asset to your entire team.

Final Thoughts

Building and sharing a published data source on Tableau Server is a foundational practice for any data-driven organization. It turns scattered, raw information into a centralized, reliable asset that enforces consistency, improves dashboard performance, and empowers everyone on your team to build reports from a single source of truth.

While mastering Tableau's workflow is incredibly powerful for teams with dedicated data analysts, we recognize that not everyone has the time to learn the intricacies of a traditional BI tool. Setting up pipelines, managing publish permissions, and remembering which extract needs refreshing can still create barriers. That's why we built Graphed. Our platform connects directly to all your data sources - from Google Analytics to Salesforce - and allows anyone on your team to create dashboards and reports by simply describing what they need in plain English. Instead of going through multiple steps to publish a curated source, you just ask questions and get instant, real-time answers and visualizations based on all your connected data.