How can I Create Data Driven Alerts using Tableau Server?
How can I Create Data Driven Alerts using Tableau Server?

During a recent consulting project, I was asked whether Tableau can alert people based on their data.

The answer was yes it can. You can use the alerts feature built into Tableau server.

These are really easy to set up and can be really helpful for driving users to Tableau server and initiating action based on the data.

Imagine sales have dipped relative to your target or call waiting times have exceeded acceptable levels. You can use data alerts to bring attention whatever is important to your business without having to constantly track data.

How can I Create Data Driven Alerts using Tableau Server?

In this example, we were looking at a test result and whether it was over or under a certain limit.

Firstly, lets look at our data:

Here we can see we have a row id, a couple of different products, a value and a limit – one limit for product A and one for product B.

Lets create a viz for this data in Tableau:

This basic bar chart with reference line shows, for each Row ID, the value as the bar and the reference line as the target.

We can publish this to our Tableau Server and then create our alert:

Once the dashboard or worksheet is published, you can navigate to the Alert menu, as seen in the image below:

To create an alert, you must first click on a numeric axis of a chart, this is to allow Tableau server to compare a value against and if a condition is met, an alert will be triggered – this is the clever bit.

Click on the numeric axis and then the alerts option on the menu ribbon.

We are then presented with the next screen and the ability to configure our alert:

Here we can see we can set a condition, like if the Value is above, equal to or below a certain value.

We can edit the subject of the alert, how often it will be sent and to who it will be sent.

These are all great options, but what if we want to be a bit more dynamic?

As we can see from the basic viz, there is a limit for each of the two products. For product A, the limit is 500 and for product B it is 700 but our alert threshold in Tableau only allows one entry. We could create multiple alerts but that might get confusing as you would have alerts triggered where the value was over one threshold but not another or under both. Not ideal.

To overcome this, as Tableau needs a numeric axis to create an alert, you can create a ratio of the value versus the target. Using this as an additional axis and the alert driver will allow different limits to be catered for.

To do this, let’s go back into the Tableau desktop workbook and create a simple ratio of value as a proportion of the limit. If this value is above 1.0, then the value has exceeded the limit and will work across any number of different limits for products.

As you can see, I have created this calculated field, added it to the columns shelf, added itself as the colour and specified a centre point of 1 on the colour so that any value above 1 is the darker orange and those below are darker blue.

I also added a reference line to show the value of 1, again to highlight those rows where the value is greater than the limit.

Now, we can publish this back to Tableau server and create our alert based on the ratio axis, as below:

I can now create my alert based on the ratio which will clearly show me for each different product an issue where the value has exceeded the limit or where it has not.

I set my condition to Above 1 and amended my Subject for dramatic effect.

I can add all the relevant people who need to be notified and voila – more flexible alerts using Tableau, a basic ratio and Tableau server.

About The Information Lab Ireland

The Information Lab Ireland is at the forefront of creating a data-driven culture in Ireland.

As part of its vision, The Information Lab Ireland regularly hosts free events throughout the country to show how being data-driven can improve decision making and lead to a better understanding of the world around us. For more on these events please like us on Facebook, follow us on Twitter or catch up with us on LinkedIn