May 20, 2024

Keeping Tableau Dashboards Reliable with Bigeye

Reliable Tableau dashboards are essential for business users to make data-informed decisions.

Christian MacDonald

Reliable Tableau dashboards are essential for business users to make data-informed decisions. If the data isn't trustworthy, it defeats the purpose of the dashboards. Let’s dive into how Bigeye helps maintain the reliability of your Tableau dashboards.

Defining Reliability in Analytics

Reliability, according to Google, is "the quality of being trustworthy or of performing consistently well." In analytics, reliability doesn’t mean being 100% correct all the time. Instead, it means minimizing data quality problems, resolving them quickly, and ensuring that end-users trust the data they see. Communication with your dashboard viewers is crucial. They should never have to guess if the data is accurate.

Common Scenario: Losing Trust in Analytics

Consider a common scenario: At 8 a.m., the VP of operations notices something off with their dashboard and contacts the data team. The data team starts investigating, but by 2:45 p.m., the VP still hasn’t received an update and has to cancel a meeting with the CEO due to unreliable data. This situation leads to a significant loss of trust. The business team shouldn’t be the ones identifying data issues. They need to trust that the data team is monitoring and addressing problems proactively.

Bigeye’s Approach: Dependency-Driven Monitoring

Bigeye tackles this issue with a different approach, focusing on the end data assets—your Tableau dashboards—and deploying monitoring backward from these assets. This is called dependency-driven monitoring. It empowers analysts and data scientists by linking their Tableau dashboards with the tables directly used and everything upstream of those tables. By monitoring everything upstream, we help data teams resolve issues faster.

In-Depth Demo: Keeping Tableau Dashboards Reliable

Let's move to an actual demo to show how this works in practice. For this demo, imagine you're part of an online resale company for albums, called Tunes of Our Youth. You have a Tableau dashboard showing the number of orders per day, average price per unit, and payment methods.

As the VP of operations, while looking at your Tableau dashboard, you notice a significant drop in sales data. This sudden drop is concerning because you need to report on your sales for yesterday. To investigate this anomaly, you open the Bigeye browser extension integrated with Tableau. This extension immediately provides an overview of the data quality and pipeline freshness, indicating whether the data is trustworthy.

The extension flags the CEO Dashboard as "at risk" due to data quality issues. It also shows the pipeline status, indicating whether the data is stale or updated on schedule. With this information, you know there is a problem, but you need more details.

To dig deeper, you click on the "View in Bigeye" button, which takes you to the detailed lineage view in Bigeye. This view shows the entire lineage of the CEO Dashboard, including all upstream tables and columns feeding into the dashboard. Here, you can see that there are 15 open issues affecting the dashboard, and the extension highlights the types of issues, such as unexpected null values and value distributions.

As you explore the lineage in Bigeye, you notice two primary issues: an anomaly in the order ID and a drop in the average price per unit in the transactions table. To diagnose these issues, you click on the transactions table and view the detailed metrics. The metrics reveal that the average price per unit has dropped from the usual $14-$15 range to $8, with a minimum dropping to zero. This indicates a significant problem.

You then go to the debug tab to run a debug query. This query shows the exact rows causing the anomaly, making it easier to diagnose the issue. If you are responsible for the entire pipeline, you would take the next steps to resolve the problem. However, if you are only responsible for creating the Tableau dashboard, you might assign the issue to the data engineer responsible for the Postgres database, informing them of the problem's impact on the CEO Dashboard.

Throughout this process, Bigeye keeps the business team informed. As issues are addressed and their status is updated in Bigeye, these updates are reflected in the browser extension. This way, the business users can see real-time updates on issue resolution, ensuring they stay informed without constant communication with the data team.

Now, let's look at how to set up monitoring for a new dashboard. Imagine you have a brand new CEO Dashboard in Bigeye that has not been monitored before. You go to Bigeye and select "Deploy metrics to dashboard." You use Bigeye's auto metrics recommendations to deploy monitoring on relevant columns and tables. This process involves backfilling 21 days of data to enable immediate anomaly detection.

You assign the metrics to yourself or another team member and add them to a collection named "CEO Dashboard." Collections allow for streamlined notifications, ensuring that alerts for this dashboard are sent to the appropriate Slack channel, email address, or Teams channel. Once you hit deploy, Bigeye begins monitoring all relevant data upstream of the CEO Dashboard, providing immediate anomaly detection and alerts.

In summary, Bigeye's dependency-driven monitoring approach ensures comprehensive coverage and keeps business users informed. By monitoring everything upstream of the Tableau dashboards and providing real-time updates on data quality issues, Bigeye helps maintain the reliability of your Tableau dashboards.

Frequently Asked Questions

1. Can we monitor Tableau's calculated fields?

One of the questions we get very commonly for Tableau is whether Bigeye can monitor Tableau's calculated fields. To clarify, a calculated field in Tableau allows you to make further transformations to your data by performing calculations within a visualization. For example, you can create a calculated field for revenue and then use that field to generate visualizations.

Currently, Bigeye does not support monitoring these calculated fields. As a best practice in the industry, we focus on monitoring the data within the conformed schemas used for the Tableau dashboards. This means we monitor the data in your databases, such as Snowflake, Databricks, or whatever data warehouse you use, rather than the custom calculations performed in Tableau.

By keeping all your business logic and important fields, like revenue, within your data warehouse, you can ensure consistency across all your dashboards. This approach prevents different recalculations in various places, providing a reliable data foundation for all your Tableau visualizations. Therefore, Bigeye monitors the data in the actual databases themselves, not the custom calculations done within Tableau.

2. How often is Tableau lineage updated?

Tableau lineage is updated on a schedule that makes sense for the customer, typically every 24 to 48 hours.

3. What other BI tools do you support?

We currently support Tableau and Power BI, with many more integrations on the way!

To see a video walkthrough, watch our webinar replay below:

Want to see a personalized demo instead? Book one here.

share this episode
Monthly cost ($)
Number of resources
Time (months)
Total cost ($)
Software/Data engineer
Data analyst
Business analyst
Data/product manager
Total cost
Common needs
Data engineers
Overall data flow. Data is fresh and operating at full volume. Jobs are always running, so data outages don't impact downstream systems.
Freshness + volume
Schema change detection
Lineage monitoring
Data scientists
Specific datasets in great detail. Looking for outliers, duplication, and other—sometimes subtle—issues that could affect their analysis or machine learning models.
Freshness monitoringCompleteness monitoringDuplicate detectionOutlier detectionDistribution shift detectionDimensional slicing and dicing
Analytics engineers
Rapidly testing the changes they’re making within the data model. Move fast and not break things—without spending hours writing tons of pipeline tests.
Lineage monitoringETL blue/green testing
Business intelligence analysts
The business impact of data. Understand where they should spend their time digging in, and when they have a red herring caused by a data pipeline problem.
Integration with analytics toolsAnomaly detectionCustom business metricsDimensional slicing and dicing
Other stakeholders
Data reliability. Customers and stakeholders don’t want data issues to bog them down, delay deadlines, or provide inaccurate information.
Integration with analytics toolsReporting and insights

Join the Bigeye Newsletter

1x per month. Get the latest in data observability right in your inbox.