What is Flow in Tableau Server?

A Tableau Flow is your personal ETL (Extract, Transform, Load) assistant, automating the repetitive, manual tasks of cleaning and preparing your data. Instead of spending hours in spreadsheets every week cleaning up messy data before you can even start your analysis, a Flow handles it for you. This article explains what Tableau Flows are, how they work on Tableau Server, and how you can use them to get straight to the insights.

First, What is Tableau Prep Builder?

Before we can talk about Flows on Tableau Server, we have to start with the tool you use to create them: Tableau Prep Builder. Think of it as a visual workshop for your data. It's a separate desktop application from Tableau Desktop (the one you use to build dashboards).

In Tableau Prep Builder, you don't write code or complex formulas. Instead, you connect to your data sources - like spreadsheets, databases, or cloud applications - and create a visual, step-by-step map of the cleaning and shaping process. You can see your data change at every step, making it easy to spot errors and understand what’s happening. The map you create in Tableau Prep Builder is called a "Flow."

So, What Exactly is a Tableau Flow?

A Tableau Flow is the saved output from Tableau Prep Builder. It's not a dataset or a dashboard, it's a recipe that details every single step required to take your raw, messy data and transform it into a clean, analysis-ready format.

Imagine you're trying to bake a cake. The Flow is your cookbook recipe. It tells you exactly which ingredients to use (your raw data sources), how to mix them (joins and unions), what to filter out (removing null values or errors), and how to prepare them for the oven (pivoting or aggregating). On its own, the recipe is just a set of instructions. To get the cake, you have to actually follow the recipe and bake it - and that’s where Tableau Server comes in.

When you publish a Flow to Tableau Server, you're giving the server your "recipe" and telling it to run that process automatically, either on-demand or on a recurring schedule.

The Relationship: Prep Builder vs. Flow vs. Server

It can be confusing to keep these three components straight. Here's a simple way to remember their roles:

• Tableau Prep Builder: The Design Studio. This is the desktop application where you visually build and design your data preparation steps. It's your interactive workspace for creating the Flow.
• Tableau Flow: The Blueprint. This is the .tfl or .tflx file you save from Tableau Prep Builder. It contains all the instructions for cleaning, shaping, and combining your data.
• Tableau Server: The Automation Engine. This is the centralized platform where you publish, schedule, and execute your Flow. It does the heavy lifting, running your blueprint on a schedule so you always have fresh, clean data ready for analysis in your dashboards.

Key Components of a Tableau Flow

A typical Flow is made up of several types of steps, each with a specific job. You add and connect these steps visually in Tableau Prep Builder to create your data pipeline.

1. Input Step

This is where it all begins. The input step is where you connect to your raw data. You can connect to a wide array of sources, including:

• Excel spreadsheets and CSV files
• Databases like SQL Server, PostgreSQL, or Oracle
• Cloud data sources like Google Analytics, Salesforce, or cloud warehouses like Snowflake and BigQuery

You can even bring multiple input steps into a single flow to combine data from entirely different platforms.

2. Cleaning Steps

This is the heart of any data prep process. Cleaning steps are focused on fixing errors, standardizing values, and making your data consistent. Common actions include:

• Filtering: Removing rows that don't meet certain criteria (e.g., filtering out test orders or test accounts).
• Removing Unwanted Fields: Getting rid of columns you don't need for your analysis.
• Splitting Columns: Breaking one column into multiple (e.g., splitting a "Full Name" column into "First Name" and "Last Name").
• Changing Data Types: Ensuring a 'Date' field is recognized as a date, not as plain text.
• Standardizing Values: Grouping similar values together (e.g., changing "USA", "U.S.A.", and "United States" all into a single "USA").

3. Transformation Steps

Beyond simple cleaning, you often need to restructure your data to make it usable for visualization. Transformation steps help you reshape your dataset.

• Aggregate: Condensing your data to a higher level of detail. For example, rolling up thousands of individual sales transactions into total monthly sales per product category.
• Pivot: Shifting your data from a wide format to a tall format, or vice versa. Useful if you have data where each month is a separate column ("Jan Sales," "Feb Sales," etc.) and you want one "Month" column and one "Sales" column.

4. Join and Union Steps

These steps are essential for combining data from different sources.

• Join: Imagine putting two tables side-by-side and connecting them based on a common field. For instance, joining your sales data (which has a CustomerID) with your customer data (which also has a CustomerID) to bring in customer details like name and location.
• Union: Think of stacking two tables on top of each other. This is useful when you have data with similar columns but from different sources. For example, combining marketing lead lists from Facebook Ads, Google Ads, and LinkedIn into a single unified list.

5. Output Step

The final step in your Flow defines where the cleaned-up data goes. The output can be one of three things:

• File: A .csv or .tde/.hyper (Tableau data extract) file saved to your computer or a network drive.
• Published Data Source: This is the most common use case. The output file is directly published to Tableau Server, ready for anyone with permission to connect to it and build dashboards.
• Database Table: Write the clean data back to a table in a database.

How to Publish and Schedule a Flow on Tableau Server

Creating a Flow in Tableau Prep Builder is only half the battle. The true power of Flows is unlocked when you automate them on Tableau Server.

Publishing Your Flow

Once you are happy with your Flow in Tableau Prep Builder, publishing it is straightforward.

1. Click the Server menu at the top of the application.
2. Select Publish Flow. If you are not already signed in, you will be prompted to connect to your Tableau Server.
3. A dialog box will appear. Here, you'll choose the Project on the server where you want to store the Flow.
4. Give your Flow a descriptive name and add tags to make it searchable.
5. Choose how to handle credentials for connecting to your data sources. You can either embed the credentials in the connection or prompt the user, depending on your organization's security policies.
6. Click Publish. Your Flow, along with its connections, is now uploaded to Tableau Server.

Scheduling Your Flow on Tableau Server

Now that your Flow is on the server, you can tell it when to run. This automates your reporting and ensures your dashboards are always powered by up-to-date data.

1. Log in to your Tableau Server and navigate to the Project where you published your Flow.
2. Click on the Flow to open its overview page.
3. Go to the Scheduled Tasks tab.
4. Click New Task.
5. From the dropdown, select one of the predefined schedules set up by your Tableau Server administrator (e.g., "Daily at 7 AM," "Weekly on Sunday," "Hourly").
6. Once you select a schedule, the task is set. Your Flow will now run automatically at the specified time, refreshing your output data source for everyone who uses it.

You can also run a Flow manually at any time by navigating to it on the Server and choosing "Run Now." This is useful for testing or getting a quick data update outside the normal schedule.

Practical Use Cases for Tableau Flows

What can you actually do with all this? Here are a few real-world examples:

• Unified Marketing Analytics: A marketer needs to create a single dashboard showing total ad spend versus revenue. The ad spend data is spread across spreadsheets from Facebook Ads and Google Ads, while the revenue is in Google Analytics. They can build a Flow that unions the Facebook and Google Ads spreadsheets, cleans up the campaign names to be consistent, then joins the result with transaction data from Google Analytics on a common date field. The Flow is scheduled to run daily at 6 AM, ensuring the spend vs. revenue dashboard is always fresh for the morning meeting.
• Full-Funnel Sales Reporting: A sales operations manager wants to track conversion rates from initial lead to closed deal. Leads come from a HubSpot export, while opportunities and deals are tracked in Salesforce. A Flow can pull from both data sources, clean and standardize the lead source fields, and combine them to create a single "funnel" output, which becomes the source of truth for all sales pipeline and conversion reporting.
• E-commerce Daily Performance Snapshot: A Shopify store owner wants a daily report combining sales, inventory, and website traffic. They can use a Flow to connect to their Shopify sales data, an inventory management system (via a database connection), and Google Analytics. The Flow aggregates daily sales, fetches current inventory levels, and pulls key traffic metrics. It then outputs a single, pre-calculated data source to Tableau Server, powering a simple performance "cockpit" that updates every morning.

Final Thoughts

Integrating Tableau Flows into your workflow on Tableau Server is a powerful way to automate repetitive data preparation. This process shifts your effort away from the tedious, manual grind of cleaning spreadsheets and allows you to focus on analyzing clean, trusted data and finding valuable insights.

At Graphed, we're dedicated to making data analysis even more intuitive. While tools like Tableau Prep offer a visual way to build data pipelines, we've designed our platform so you don't have to build them at all. Just connect your marketing and sales platforms - like Google Analytics, Shopify, or Salesforce - in a few clicks, and then ask questions in plain English. Instead of manually cleaning and joining data, you can simply ask, “Show me my ad spend from Facebook and Google versus my Shopify revenue this month” and get a live, automated dashboard in seconds, letting you skip straight to the answers you need about what's working with Graphed.