Skip to main content

Tableau Data Driven Alerts

Data driven alerts are a built-in feature in Tableau that allow you to tell Tableau Server to notify certain people if your data reaches or surpasses a set threshold. They can be set up to alert via email or a notification within Tableau. 

Data-Driven Alerts as Part of Requirements

It is recommended to collect requirements on potential alerts as part of the requirements gathering process. 

Sample questions to ask include:

  • If there is an issue, what will the metric look like?
  • When should we take corrective action? 
  • Who is responsible for acting when the alert triggers?

Setting Up Data-Driven Alerts

To set up data driven alerts,

  1. Select a continuous numeric axis of a chart that is not a Gantt chart or map. 
  2. Select watch from the Alerts side panel and choose alerts. 
  3. A dialogue box will then pop up, allowing you to edit the alert. You can choose the trigger condition (above or equal to, above, below or equal to, below, or equal to), threshold, the subject line of the notification and how often users should receive the alert. 

     

    Helpful Note:

    When selecting the threshold, a pink line will temporarily appear, allowing you to visualize where the threshold is in comparison to the current data. If you want others to be able to see the alert and add themselves to it, select Make visible to others. Other users will then be able to see the alert frequencies and add themselves to the alert by selecting Add Me.

Additional Considerations when Setting up Data-Driven Alerts

  • Subset of Data
    • If you want to create an alert for only a subsection of the data, you need an axis that is specific to that subset of data. This can be done by creating a view for that subset of data, which can be accomplished through command buttons that allow you to keep only or exclude. 
  • Data Overload
    • Be cognizant of data overload. If there are too many false alarms (i.e., action is not required), then users may ignore the alert when action is required. 
    • During future iterations, it is important to follow up with stakeholders around data-driven alerts. 
  • Recent Data
    • A good rule when trying to identify data-driven alerts is to see if there is historical data. From there, it is recommended to pull daily results from the past 90 days and identify what is 2 standard deviations from the daily average. 
    • If you do not have historical data, a subscription may be a better use than data-driven alerts. 

Managing Alerts

You can manage alerts from the My Content area of the Tableau web pages or from the alert emails. Below the alert emails are links letting you manage the alert or add or remove yourself from it. The alert owner can change the alert settings through the Edit this alert link. 

Alert Failures or Suspensions

Alerts can fail because of issues such as temporary connective issues (which will be fixed on its own), if a data source is removed, if credentials to data have expired, or the worksheet or sheet that the alert was created on is removed. If an alert fails, the alert owner will receive an email stating that the alert fails with a link to fix the problem. You’ll then receive an email when the alert is working again. 

If an alert fails too many times it will be suspended. The alert owner/administrator can resume a suspended alert through the My Content area of the Tableau web pages or by clicking Resume Alert in the notification email.