Creating complex dashboards using Tableau
Tableau gives us the capability to create complex and useful dashboards which can be very helpful to look at data from various angles. As part of the Dashboards and Storytelling with Data , course on Coursera I have learnt how to create a dashboard using the Sales SuperStore data set which is available for download here. I want to give a heads up that this is a lengthy article with tons of screenshots which should help you follow along the process. At at end of the article, I have added a cool video to demonstrate the power of the dashboard we have created. The process is lengthy and should take close to 1 hour of your time but it is also extremely rewarding to look at the beautiful visualization I have created as part of the article. Lets go!
Before we get into the actual process of creating a visualization lets review some key best practices in creating dashboards.
- We want to choose the metrics that matter.
- We want to keep it visual.
- We want to make it interactive for collaboration.
- We want to pull data from all sources to get the full picture and keep the dashboards up to date or refreshed. And, finally,
- we want to make it simple to access and easy to use.
Now, lets think about a few important considerations for creating dashboards
- Just because something looks cool, it does not mean that it has to be on the dashboard. As part of your job as a data analyst you might create hundreds of worksheets to perform exploratory analysis but only 5 of them could end up on the dashboard.
- Our second consideration is to make it accessible. We need to tap into the iconic memory of our users to help them walk away with key data and trends.
- Our third consideration is the purpose of the dashboard. Who is the user and what is their role?. How will they be viewing the dashboard and what will they do with the dashboard? What value will the dashboard bring to them? Will it help management deal with what is important?Is the dashboard specific or organization wide? Will it be historical, real-time or snapshot
The course goes through all these aspects in detail and they are very important to think about before creating a dashboard.
Customer Scatterplot
Lets create a scatter plot with Sales in Columns and Profit in rows. To see the entire scatterplot , click on Analysis and un check aggregate measures.

Drag profit ratio onto color which creates a beautiful visualization

Click on Shape and select the Circle and then click on Color and select Edit Colors. The color scheme in our scatterplot is a little imbalanced. Lets change that. Click on Advanced, check the Start option and change the Start value to -0.5. This way we have a balanced color scheme range of -0.5 to 0.5


A first glance at our visualization tells us that Profit and Sales go hand in hand but we need to dig deeper to find out. Lets move the Customer Name and Region fields to Detail so we can edit the Tool Tip. If you are not already familiar with Tool Tip, it helps declutter visualizations so that we can show necessary information when users hover over the visualization. Click on Tooltip and modify it like in the picture below so that the customer’s name is emphasized. I increased the customer name size to 15 and made it bold. This is how it shows up on the scatter plot. Finally, lets rename the sheet to Customer Scatterplot.

Customer Rank
We will create our next worksheet and name it Customer Rank. Drag Sales over to columns and Customer Name to rows. Here we see every customer with a bar indicating how their Sales are doing.

Let sort the data by clicking on the downarrow on Customer Name in rows and select the descending sort order, sort by Sales and aggregation as Sum. This gives us a sorted list of customers and their Sales.


Lets add Profit Ration to Color and edit the colors to choose a range of -0.5 to 0.5. This has been explained in the previous visualization. This highlights the customers with the least profit ratio of -50% in red. Drag the profit measure to tool tip and Sales to Detail. To be able to see the rank of a customer in the tool tip, click on the down arrow on Sales, select quick table calculation and then select rank.

This shows the rank of each customer in the tool tip when you hover over the chart, this lets users access additional information from the visualization when they need it. Format the text in the tool tip like in the previous visualization to make it look like this

Customer overview
Moving on to the next visualization, open a new sheet and name it Customer Overview. Move Region to rows and Profit Ratio to color and edit the colors by choosing the range of -0.5 to 0.5 . This shows us that each region has some profit but it doesn’t tell us a lot.

Drag customer name over to Detail and select Measure Count Distinct from the down arrow. This should change the visualization to the same as above. Drag Sales, quantity , profit and profit ratio to details . Next step is to create a calculated field from the Sales measure.

Once we are done creating the calculated field, drag the measure name and measure values over to columns and keep Sales per Customer, Sales, Quantity, Profit and Profit Ratio in measure values and remove everything else in measure values to make the visualization less cluttered. Next step is to remove the horizontal axis and add check the measure values box in tool tip. This is what we have until now,

We have a lot of information in the tool tip which makes it kind of confusing. Lets edit the tool tip

This is what we have after editing the tool tip.

Dashboard iteration -1
We can now create a dashboard using the three visualizations we have created. Click on the dashboard sheet and rename it as Customer Dashboard. Drag and drop all three sheets onto the dashboard and move them around until you can see something similar to this image

We now have a basic dashboard but it is not interactive. We can add interactivity to a dashboard by creating filters. Filters give our audience the ability to navigate our dashboard. A quick look at our dashboard tells us that there is no time indicator on our dashboard. Lets create a time filter by creating a new sheet and drop order date on rows which tells that there are 4 years 2014–2017. We need our audience to be able to look at information year by year . To be able to have the filter in the dashboard , Tableau needs it to be in the worksheet. In the Customer Scatterplot sheet, drag Order Date to the filters section and select years from the pop up window.


OK, so we created one filter, two more to go. Our next two filters are category and segment. On the same worksheet, Customer Scatterplot add category to filters and select Use all. Repeat the same process with segment. We have 3 filters now. To add these filters to the dashboard, click on the tiny arrow on the Customer Scatterplot pane in the dashboard and select each of the filters one by one from the filters option.

Dashboard iteration -2
Next step is to edit the Year of Order Date filter name to Year because that is simple to understand. You can do this by double clicking on the filter name. Click on the down arrow for each filter and select Single Value (drop down) or Single Value ( list). Single Value drop down saves us space and Single Value list displays all the possible values in the filter.

If you play around the Year filter you should see that it only changes the data in Customer Scatterplot but we need the filter to be applied across the dashboard. To fix this, click on the down arrow in the year filter, select Apply to Worksheets and then Selected Worksheets.

This should open up a dialogue box in which you should select all the worksheets. All on the dashboard and None buttons on the dialogue box are quicker ways to select all or none. If you now change the year from the filter, the entire dashboard changes. Repeat the same steps for the Category and Segment filters so that the filter applies to all sheets on the dashboard.
The next step is to select only the relevant value , to view values only based on the filter context. To do this click on the down arrow on the filter and select the ‘Only Relevant Values’ option. Do the same for segment as well. This makes it much accessible for our users to explore the dashboard.
Dashboard iteration -3
Some issues we see with our dashboard so far is that the Customer Ranking data is hard to access as it goes far down the page. Our users need to be able to connect the Scatter plot data with Customer Ranking which they are not able to right now. We can fix this by creating a dashboard action. The way to do it is to click on Dashboard from the top menu and select Action. This opens up the Actions box pop up.

Click on the Add Action button and then click on filter. This should open up an other pop up window. Follow along with me here by selecting the options as in this picture

Now, click on Add Filter which opens up an other pop up window. Select Region from the Field drop down and click OK.

Click OK on the other pop up windows. You can verify, if this is working by clicking on any bar from the Customer over view. This should filter the entire dashboard and show values only for that geography.
Dashboard iteration-4

We can also have this work for each Customer by adding an other Dashboard Action and then click on the Add Action button and select highlight.

Click OK and now we have two actions, Filter region and Highlight Customer. When we hover over any customer from the scatter plot, the dashboard displays filtered data for that customer which is really cool.

Something else we can do is to change the size of the dashboard on the left pane to Automatic. We still have a problem with Customer Ranking being really big. We will add an other filter action which should solve this problem.

Add Customer Name to target filters and click OK on all pop ups. Now when we click on any customer from the dashboard the corresponding ranking value come up.
Final Dashboard

We have actually created a very powerful dashboard which allows our users to look and iteratively filter down to information they need. Watch the dashboard in action in the video.
Congratulations, countless screenshots later, we have successfully created a complex dashboard which helps our users look at data and filter deep into it.