UX review: Tableau Online
How easy is it to get started with Tableau Online, the web app edition of one of the most widely used data visualization tools? Here’s my own experience after signing up.
Tableau is the most widely-mentioned data visualization tool. It’s mentioned by all the seven “Best Data Visualization Tools” articles that I found as top google search hits for “data visualization tools” over the past one year — a runner-up is Power BI, mentioned by six of the seven articles. (See the reference at the bottom of this article for the list of these seven articles — I excluded the articles by data visualization tool producers themselves: Visme and Chartio, even though their articles came high in the search results).
Today I signed up for its cloud app, Tableau Online, to see how easy it can be used. Here’s my “user experience”.
Interactive tutorial
Once you log in for the first time, you’re immediately encouraged to start an interactive tour of how to use Tableau Online. You can dismiss it by clicking X, if you’re already an experienced user.
So I follow the instruction. By clicking “Create”, the instruction disappears. I manage to remember what to do: “select Workbook”. But ideally the instruction should stay there when the user clicks the actual UI elements.
Then I’m taken to this screen:
I follow the instruction: click “Superstore Datasource”, which removes the instruction but turns the bottom-right “Connect” button into a blue button. That’s nice. It naturally makes me click the “Connect” button.
Then I’m taken to this screen:
As I have a bit of knowledge about data structure, I interpret dragging to the “Rows” shelf (outlined in light blue, which is nice) as specifying the unit of observation. But the instruction says “Drag the Sales field”, which should be one of attributes…
Anyway, I follow the instruction, which leads me to this screen:
The Rows field now says “SUM(Sales)”. Seems like the sales figures are averaged across all observations. Some explanation would be nice here.
I continue following the instruction: drag the Order Data field to the Columns shelf, which takes me to this screen:
I want to see how the line chart looks like, but I’m scared to remove the instruction panel by clicking X, because I expect I won’t be able to come back to the interactive tour. I tried to drag the blue box to see the chart, but I couldn’t drag it. It just stays there.
I give up and follow the instruction: drag the Segment field to Color on the Marks card, which leads me to this screen:
Now the sales figures are apparently averaged by segment, producing three series in the line chart, and I’m congratulated.
It’s a nice interactive tutorial. I do get the very basic way of creating a line chart with multiple series over time.
But I’m lost. How can I make a different kind of chart, say, bar charts?
Also, the tutorial would be much nicer if I were told what kind of chart I was going to create. We never ever create a chart without knowing what to show. In this sense, this interactive tutorial doesn’t live up to my expectation.
Anyway, let’s follow the instruction. I click Next, expecting the second interactive tutorial to begin. But I see this:
Back to the beginning. No instruction whatsoever. I’m lost.
Uploading my own data
Maybe I should follow the same steps: click “Create” and select “Workbook”, which takes me to this:
I’m lost. How can I upload my own dataset? After several seconds of wondering, I tried clicking “Files”:
Yes, I can now upload a dataset. Choosing the data file in my computer leads me to:
The file name shows up (“Freedom_House_2013–2021”). This data file is not cleaned. I wonder if Tableau takes the tedious role of data cleaning… It seems it can. I click the checkbox labeled “Use Data Interpreter”, which “may be able to clean your Microsoft Excel workbook”. Then I see this:
It’s not clear whether data cleaning was successful or not. There’s no feedback, breaching the basic principle of user interface design…
Move on. I click FIW13–21, which is the name of a worksheet containing the data that I want to visualize. Now I see the data:
Wow, the data cleaning was indeed successful. In the original Excel worksheet, the top row is a merged cell with the table title text. That’s automatically removed.
Now I drag this worksheet to the main panel, and I see this:
I see the Update Now button. I don’t understand what it means. But seems like it’s the button I should click now. So I click, and I see this:
Now the data is shown in the bottom panel.
Seems like it’s ready to be visualized. But where should I click…
Maybe I should save this data. So I click File in the menu bar:
… and select Save As.
I enter the name “FreedomHouse”, and click the Save button. The blue button color is really effective. It always clearly tells where to click for moving on to the next step.
Then I’m taken to this familiar screen:
My own data visualization attempt
The Interactive Tours button shows up. But I think I can handle it. So I just repeat what I have learned earlier: drag a field on the left column to the Rows shelf and then another to the Columns shelf.
But it’s not entirely clear which should go to the Rows and which to the Columns…
I want to create a bar chart where each bar represents a country with the bar length showing the degree of lacking in political freedom. By the way, the data I’ve uploaded is Freedom In the World, the annual cross-country panel data on political freedom from 2013 to 2021.
Each row in the data is country-year, identified with two columns: Country/Territory and Edition (referring to the year of observation). It’s a composite primary key.
So I drag Country/Territory to the Rows shelf:
And also drag the “Edition” field to the Rows shelf as well, expecting that the Rows shelf is for primary key:
Oops, the years are averaged by each country/territory…
I undo it by pressing cmd + z, which works!
Now I drag the PR rating field (which measures the degree of political freedom) to the Columns shelf:
This time I’ve managed to create a bar chart, although the political freedom score is summed up across years, rather than averaged. It’s not clear how to select the method of aggregation…
Just to show off that I still remember what I learned from the initial interactive tutorial, I drag and drop the “Region” field to the “Color” box:
Now each bar’s color represents which region a country belongs to.
I want to sort the bars by its size. But I have no clue… Seems like this is the end of what a person can do with Tableau Online without referring to the documentation or more thorough tutorials.
Verdict
Here are what I think is good, bad, and ugly about Tableau Online.
The Good
A chart can easily be created just by dragging the data fields to where they should be, though it’s not entirely clear which fields should be dragged to where.
The blue button always tells me where to click to move on to next stage.
A simple data cleaning can be done with just one click.
A chart type is automatically chosen given which data fields the user chooses, though it’s not entirely clear how to choose a chart type when the user has a clear idea of what chart type to use.
The Bad
The meaning of Columns and Rows is not clear to me at all.
The interactive tutorial is not always very helpful: it didn’t tell me how to upload data, for example. I was just lucky to figure it out on my own.
The Ugly
I cannot do more than creating a line chart or a bar chart, without taking a tutorial or reading the documentation seriously.
The above verdict only concerns the usability for casual users who just want to quickly visualize their data.
Of course, it can be a different story if I take time to learn how to use it more seriously. Otherwise Tableau Software, the producer of Tableau, wouldn’t have been purchased by Salesforce for 15.7 billion dollars in 2019.
References
Neil Patel. “Best Data Visualization Tools”, neilpatel.com, undated.
Andrea Laura. “Top 10 Data Visualization Tools for Every Data Scientist”, KDnuggets, May 2020.
Bernard Marr. “The 9 Best Analytics Tools For Data Visualization Available Today”, Forbes, May 22, 2020.
Tim Stobierski. “TOP DATA VISUALIZATION TOOLS FOR BUSINESS PROFESSIONALS”, Harvard Business School Online, Jan. 12, 2021.
Anurag. “Top 10 Big Data Visualization Tools to Make Sense from Clustered Data”, newgenapps.com, Apr. 18, 2020.
Glorify. “5 Best Data Visualization Tools of 2021”, glorifyapp.com, Oct. 31, 2020.
Octoparse. “Top 30 Data Visualization Tools in 2021”, Octoparse Blog, Jan. 20, 2021.
