How to Create an Extract in Tableau Public

Building a compelling visualization in Tableau Public is a great first step, but what happens when you try to publish it and the performance grinds to a halt? Or worse, how do you share your dashboard when the underlying data is a huge file sitting on your desktop? The answer lies in creating a Tableau Extract. This article will walk you through exactly what an extract is, why it's essential for Tableau Public users, and how to create one step-by-step.

First, What Is a Tableau Extract (and Why Should You Care)?

When you connect to data in Tableau, you have two options for your connection type: Live or Extract. Understanding the difference is crucial for building fast, portable, and shareable dashboards.

A live connection, as the name suggests, maintains a direct link to your original data source. When you drag a field onto your canvas or apply a filter, Tableau sends a query directly to that source - whether it's an Excel file, a Google Sheet, or a SQL database - and waits for a response. For small datasets, this is fine, but for larger files or complex visualizations, it can lead to slow load times and a frustrating user experience.

An extract, on the other hand, takes a different approach. When you create an extract, Tableau takes a "snapshot" of your data and saves it as a highly compressed, optimized file right inside your workbook. This special file format, ending in .hyper, is engineered for speed. All of your interactions - filtering, sorting, calculating - now query this hyper-fast local copy instead of the original source. This dramatically improves performance and makes your dashboards feel quick and responsive.

For Tableau Public users, extracts aren't just a "nice-to-have" - they're often a necessity. Since dashboards on Tableau Public can't connect live to your local computer's files (like an Excel or CSV file), creating an extract is how you package that data up so it can be published and viewed online.

When to Use a Tableau Data Extract

While extracts offer big performance boosts, they aren't necessary for every project. A live connection is perfectly suitable when you're working with a small dataset or need real-time data from a cloud source Tableau Public supports. However, you should always create an extract when:

• Your dashboard is slow. This is the most common reason. If you notice a lag every time you apply a filter, pulling your data into a .hyper extract is the first thing you should do.
• You need to share your workbook. To publish a workbook that uses a local data source (like a CSV or Excel file) to Tableau Public, you must create an extract. This bundles the data with your visualization so others can view and interact with it online.
• You're working with a very large dataset. Extracts allow you to filter and aggregate your data before it's brought into Tableau, creating a smaller, more manageable subset to work with.
• You need offline access. With an extract, the data is saved in your Tableau Workbook (.twbx file). This means you can open, edit, and interact with your dashboard without needing an internet connection or access to the original file path.

Step-by-Step Guide: How to Create an Extract in Tableau Public

Ready to create your first extract? The process is straightforward once you know where to look. We'll use connecting to a simple Excel file as our example.

Step 1: Connect to Your Data Source

First, open the Tableau Public desktop application. On the "Connect" pane on the left, choose your data type. For this walkthrough, we'll select "Microsoft Excel." Navigate to your file and click "Open."

Tableau will then take you to the Data Source page. Here, you'll see a preview of your data, and you can drag sheets onto the canvas to join them. Once you have the data you need, you're ready for the most important step.

Step 2: Switch Your Connection Type from Live to Extract

Look to the top right of the Data Source page. You'll see two radio buttons under the "Connection" heading: Live and Extract. By default, Tableau will likely be set to Live.

Go ahead and click the "Extract" radio button. This action tells Tableau that you don't want to query the original Excel file anymore. Instead, you're preparing to create a specialized .hyper extract file. It's that simple.

Step 3: Edit Your Extract (Optional but Recommended)

Once you select "Extract," a new link appears next to it that says "Edit..." Clicking this opens the Extract Data dialog box, which gives you powerful options to make your extract even more efficient. While this step is optional, it's a best practice that separates beginners from seasoned users.

Data Storage

You can choose between "Logical Tables" and "Physical Tables." For most use cases, the default setting of Logical Tables is ideal. This stores the data in the same grouped manner as in the data canvas, which offers greater flexibility down the line.

Filters

This is arguably the most useful feature in the Extract dialog. Click the "Add..." button to create a filter on your extract. This filters your data before it's ever pulled into Tableau's memory, which is fantastic for performance. For example:

• If you only need to analyze data from a specific region, you can add a filter for "Region = 'West'."
• If your dataset contains ten years of historical data but your analysis is only on the past two years, filter your date field to include only the necessary range.

By preventing Tableau from even loading unnecessary data, you make the extract file smaller and your dashboard significantly faster.

Aggregate

This option allows you to pre-aggregate your measures to a specified level of detail. For instance, if you have daily sales data but you know your charts will only ever show monthly totals, you can choose "Aggregate data for visible dimensions." Tableau will then sum up the sales for each month and store only that aggregated value, dramatically reducing the number of rows in your extract.

A word of caution: This is a permanent change for this extract. Once you aggregate, you cannot drill down to the daily level unless you create a new extract without this setting. Use it only when you are certain you won't need the lower-level detail.

Number of Rows

Finally, you can choose to extract "All rows" or a sample (e.g., the "Top" 10,000 rows). Sampling your data is useful when you're first exploring a massive, multi-gigabyte dataset and want to quickly build a prototype without waiting for the full extract to create.

Step 4: Generate the Extract File

You've connected to your data, selected the "Extract" option, and customized its settings. Now it's time to generate the file.

The extract isn't actually created until Tableau needs to access the data. The easiest way to trigger this is to navigate to a worksheet. Click on "Sheet 1" at the bottom left of your screen.

A "Save Extract As" dialog box will immediately pop up. This is where you'll save your .hyper file. Name it something descriptive and save it in the same folder as your project. After a brief loading process (which varies depending on the size of your data), your extract is complete!

You can confirm you're using an extract by looking at the icon next to your data source in the "Data" pane. A single blue cylinder represents a live connection, while two cylinders with an arrow indicate you're successfully working with an extract.

How to Refresh a Tableau Extract

Since an extract is a snapshot in time, it won't update automatically if your original data file changes. For example, if someone adds new rows to the source Excel file, those changes won't immediately appear in your dashboard.

To update your data, you need to perform a refresh. This tells Tableau to go back to the original source file, pull in the latest data, and overwrite the existing .hyper file.

Here's how to do it:

1. In the "Data" pane on the left side of your worksheet, right-click on your data source.
2. In the context menu, hover over "Extract."
3. Click on "Refresh."

Tableau will briefly reconnect to your original file, process any new or changed data, and update your extract. All of your visualizations will then update accordingly. Remember, you must have access to the original data file on your computer for the refresh to work.

Final Thoughts

Mastering extracts is a fundamental skill for anyone serious about using Tableau Public. They are the key to building high-performing dashboards, especially as your data size grows, and they are essential for packaging your local data files for publication online. By following the steps to create, configure, and refresh your extracts, you can ensure your audience enjoys a smooth and interactive experience.

The manual process of creating extracts and refreshing data, common in tools like Tableau, highlights the traditional complexities of business intelligence. To streamline this workflow, we built Graphed to be an easier, smarter alternative. It eliminates manual setup by deeply understanding your data sources, allowing you to connect platforms in one click and use simple natural language to generate entire dashboards in seconds. This replaces the routine of downloading CSVs, managing extract files, and wrangling different tools just to get answers.