Your web browser is out of date. Update your browser for more security, speed and the best experience on this site.

Update your browser
CapTech Home Page

Blog July 3, 2018

Tableau Can Do That?!

Tableau is an extremely powerful tool to visualize data, not only offering a myriad of data sources to connect to, but also providing the ease and simplicity of designing dashboards. The 'Out of Box' chart types (visible when clicking on 'Show Me' in the designer window) are a great starting point, each having several benefits based on the data you wish to visualize.

Sometimes, however, you feel the need to go beyond these available chart types. Some questions that need answers are better visualized by a different chart type, not available in Tableau. We were presented with a similar challenge at one of our client recently.

This client is a big brand name in the hospitality industry with multiple properties worldwide. They wanted to visualize relationships between the various parties they engage with. Think Johnny engaging with Wes, who in turn also engages with Bill. How would you visualize this relationship? Bar charts are definitely not going to help here. Neither are heat maps or scatter plots! Network graph certainly sounds like a good option here. "But wait! I do not see that as a chart type in the 'Show Me' drop down. How can I accomplish it?"

We will be exploring these options through a four-part blog series, with the first one focused on some of the custom chart types, followed by three other parts highlighting the detailed steps to develop these custom charts.

The beauty (some call it a limitation if you can't manipulate the data) of Tableau is once you structure the data in a certain way, it opens doors to develop new visualizations; visualizations that do a better job of answering the business questions the traditional chart types may not answer.

Some of these techniques were piloted by familiar names in the tableau community viz. Stèphane Frèchette, Jeffrey Shaffer, Olivier Catherin and Bora Beran. They have done a fantastic job laying the ground work for these chart types that we as Tableau developers can use.

Let's take a look at some of these chart types

Network Charts

How many of you still remember the Graph theory from the good old school days? For those that may not, here is a quick refresher.
In mathematics, graph theory is the study of graphs, which are mathematical structures used to model pairwise relations between objects.

https://en.wikipedia.org/wiki/Graph_theory

Notice the word 'relations'. This graph is a great option for visualization relationships between different members of a dimension.

Using this theory, we can design a simple network graph in Tableau. Going back to our original problem, we can now tell Liz is engaging with several different people, while Tina is engaging with only one! Each slice of pie represents the number of people that person is engaging with.

Sankey Diagram

Sankey diagrams are a specific type of flow diagram, in which the width of the lines are shown proportionally to the flow quantity.

This is a great chart type to visualize metrics and how they change based on relationships between dimensions.

Using the global superstore dataset that comes preinstalled with Tableau, we have created a Sankey chart that shows the Percent of sales and how it changes between the different product categories and the Markets they are sold in. We can see that most of the furniture and technology sales are made in Asian markets, while Europe consumes a lot of office supplies. Europe is also a strong market for technology products.

Sunburst Diagram

Sunburst chart is typically used to highlight hierarchy in the data, via the concentric circle patterns. The granularity of the data increases as you move outwards, with the outermost circle indicating the highest level of granularity. This type of chart is used to compare relative sizes between various dimensions.

For example, using the global superstore data, we can conclude that Southern US has the highest sales, with Office supplies serving as the primary contributor.

Most of these chart types need some sort of data manipulation to get to the final product. Some of the techniques are simple, while some use advanced concepts such as data densification. However, the end result is a visualization that helps answer the business question with charts the business may not have seen before.

So, the next time you have your business user say, "No more tree maps or pie charts please!", think of these chart types and if they satisfy the use case!