How to Do Performance Testing in Tableau

A slow-loading Tableau dashboard is more than a minor annoyance, it’s a roadblock to getting the insights you need. Instead of making quick, data-driven decisions, you're stuck waiting, and your team loses trust in the tool. This guide will walk you through how to use Tableau's built-in tools to find exactly what's causing the slowdown and how to fix it.

Why Is My Tableau Dashboard Slow? Common culprits behind the lag.

Before launching into testing, it helps to understand the common issues that cause poor performance. Usually, the problem lies in one of four areas: your data, your calculations, your dashboard design, or the server environment itself.

Inefficient data connections:

• Extracts vs. a Live Connection: A live connection queries your database in real-time. If the underlying database is slow or the query is complex, your dashboard will be slow. An extract, on the other hand, is a snapshot of your data stored in Tableau's high-performance in-memory database. While extracts need to be refreshed, they almost always provide faster performance.
• Data volume: Are you trying to pull millions of rows into a dashboard that only displays aggregated data? You wouldn’t bring every scrap of product data when you can focus on data that can directly improve your business. Pulling only what you need improves speed.
• Complex Joins: Joining multiple large tables, especially with complicated conditions, can significantly slow down query performance. Oftentimes this data work is happening at the database level instead of in Tableau. A little preparation can go a long way when it comes time for analyzing information and making decisions from your company’s sales reports or financial data.

Overly complex calculations and statements:

• Row-Level vs. Aggregate: A row-level calculation (e.g., [Sales] - [Cost]) is performed for every single row in your dataset, which significantly increases processing time. A lot of time can be saved by using aggregated calculations like summing values (SUM([Sales])) instead.
• Heavy String Calculations: Functions that manipulate text, such as SPLIT or string CONTAINS, can take lots of memory and reduce runtime. These types are popular, but remember to limit their usage to avoid overtaxing resources.
• Too many LODs: Level of Detail expressions are extremely powerful. However, using them carelessly creates complex queries, leading to slow dashboards.

Complicated dashboard design:

• Too many charts: Dashboards should tell a story. More visuals do not inherently equal more value. Trying to include every KPI on a dashboard can lead to "analysis paralysis." It's best to focus on dashboard designs with a limited number of graphs for better clarity.
• Too many Marks: Plotting excessive data points, like thousands of orders, can slow down the system. Instead, use techniques like geographical hex maps to aggregate data efficiently.
• Lots of filters: Each interactive filter requires calculations, which can tax system resources. Focus on essential filters to reduce unnecessary processing.

How to find performance issues during testing

To diagnose performance issues efficiently, use Tableau's built-in Performance Recorder. This tool tracks interactions, helping to identify slow processes and improve efficiency.

Step-By-Step Instructions for Using the Performance Recorder:

1. Start the Recording: In Tableau Desktop, go to the top menu and select Help → Settings and Performance → Start Performance Recording. This activates performance recording, indicated by a check icon.
2. Run Your Dashboard Interactions: Open your dashboards and interact with them as usual. All actions are recorded, capturing data about each step, filter, and visualization.
3. Stop the Recording: Once you've completed your tests, stop the recording by going to Help → Settings and Performance → Stop Performance Recording. This generates a performance report.

The report will display detailed analytics about the dashboard’s performance, providing insights into areas needing optimization. You’ll see data visualizations that help identify bottlenecks and high-processing areas.

Analyze the Information:

With the generated performance report, dive into the data to understand specific issues. Use the timeline view to track significant processing delays. This helps pinpoint areas requiring optimization, such as reducing the number of visualizations or optimizing queries.

Final Thoughts

Using tools like Tableau’s Performance Recorder allows you to diagnose and address performance bottlenecks effectively. This empowers you to optimize your dashboards for quicker loading times, enhance user experience, and make data-driven decisions efficiently. Start today by leveraging these insights and maximizing the potential of your Tableau dashboards.

To further simplify your Tableau experience, consider integrating tools like Graphed. It can streamline your workflow and enhance collaboration in creating robust visualizations.