What Makes a Successful Tableau IronViz Visualization (from an IronViz Qualifying Judge)
2021 has been an incredible year for me professionally and personally. Still, I was super excited to be selected as a Round 1 IronViz Qualifying Judge with all of these super talented and first-rate members of our community. With the announcement of our Top-10 finalists in the Data + Joy-themed IronViz Qualifier contest, I’m ready to write about the experience and help you determine what to consider with future qualifying events.
Before we get into this post, here are some things I will not share here or privately:
- Specific visualizations reviewed
- Itemized scoring breakdowns
However, I will cover traits of a successful IronViz visualization based on the trends of high-scoring data visualizations from the 64 visualizations I reviewed and the tendencies of the ten finalists.
Iron Viz Qualifying Judging Background
Judges are expected to look at visualizations objectively. Tableau made significant efforts to remove personally identifiable information from the visualizations, and we did not use Tableau Public to find and review vizzes. Instead, the submitted visualizations were hosted elsewhere. The effort (notably Tableau’s AB Commendatore and Andrew Grinaker (Twitter)) extended to do this was no small feat, but it helped keep the scoring fair. To avoid seeing IronViz qualifying entries, I muted references to IronViz on Twitter, avoided Tableau Public, and would not engage if a person requested feedback on any Tableau visualizations during that period. It wasn't easy to do as I’m super curious and felt I missed out on a lot of excitement, but objectivity was essential. Of course, IronViz judges do not submit their entries for the contest as that would be an apparent conflict of interest. In addition, we were given the option not to judge vizzes if a visualization could not judge objectively (e.g., potential breeches above).
Judging Process
Each judge at the Round 1 level was delivered 60+ visualizations scattered among all 320 entries. Multiple judges reviewed every visualization. The highest scoring visualizations were shared with the 2nd Round judges. Every visualization was scored based on…
- Analysis of how the data supported the topic, its cleanliness, and its application to the study, including calculations
- Design or how the data worked together and served in a manner the end-user can process it visually. These questions should come to mind… Is this the proper use case for the chart? Can I quickly read the text and visuals? Do the colors have a purpose? Does the visualizing consider accessibility?
- Storytelling or how effective the visualization is at using the analysis and design to focus on the topic. Is there a little movie playing in an onlooker’s head while navigating the subject and its trail? Does it engage and enlighten more than the data and charting do on their own?
An analogy-laden breakdown of a quality IronViz entry
I’m a little hungry, so I cannot think of a better time to present this analogy. I’m no chef, but I witnessed many of them on TV — or at least enough to cobble this analogy together.
The analysis stage is shopping for ingredients with a recipe in mind. Next, it would be best to consider what ingredients, guests, and dishes are needed for the 8-course meal. Next, the design stage considers the cooking method, cuisine, timing, venue, and plating. Finally, for storytelling, the chef finds the theme, order of delivery, timing in between, and other delicate touches to add a sense of adventure and coherence to the event.
IronViz is the fine-dining of data visualization. The most successful entrants spend significant time and energy putting all of these pieces together. The cuisine, venue, and cooking methods vary from one author to the next. However, they all have these general characteristics:
- Data clarity and care — The author spent significant time collecting data for the visualization, verifying calculations, working out all the logic needed to render their design ideas, and taking out unnecessary bits. Given the topic, the best data was carefully collected or compiled.
A tip if Tableau resumes providing a dataset for qualifying entries in the future
Verify whether you can append your data or additional data to the dataset to make a more personal visualization. The more of the visualization comes from you, the less it feels like an assignment and more of a passion project. Passion projects generally yield stronger results as more care is put into the visualization to engage and push yourself to showcase your skills.
- Design efficiency and mastery — With this understanding of data, they already knew what would be on the dashboard. Many people do this by wireframing or making a pre-design of the visualization. It’s a rough idea of what will be in a final visualization and provides a course.
- Storytelling coherency and engagement — Authors step back and consider the big picture of something they are passionate about. Then, with that vigor, they want to convince you that this is something you should also feel with the least amount of steps. It’s intended to give you that a-ha moment and get you to think about it after you stopped looking at it.
- Planning, coordination, time, and flexibility — before completing this work, they think deeply about the placement anything has and how to fine-tune it. Authors may do this by taking time between steps and looking at it over days or weeks — seek feedback on the design and verify another person sees what they vision and want to communicate — be flexible enough to re-iterate and make necessary adjustments on all steps to render something intended and powerful.
3 General Trends of Top-10 IronViz Data Visualizations
I saw the following similarities with many of the higher scoring entries (in addition to the items mentioned above). In 2022, the trends may change, so you will want to see what the community is doing (particularly in Sarah Bartlett’s (Twitter | Site) #IronQuest).
- Long-form visualizations are the rage.
All visualizations that reached Top-10 required vertical scrolling — most significantly. Only Ryan Soares (Twitter | Tableau Public) had a layout that wouldn’t require scrolling from many devices (configuration of 1000 pixels width by 700 pixels length over seven dashboards).
Note: the default desktop format and a general rule of thumb dimensions for a business dashboard are ~1,400 x 800 to make it accessible without scrolling on most desktops.
Top-3 Finalist Samuel Parson (Twitter | Tableau Public) was the only finalist with a much wider than longer visualization (a 6,000-pixel width x 2,000-pixel length).
The remaining eight had median dimensions of 1,400-pixel width and 5,650-pixel length with averages of 1,324 and 6008 respectively, or approximately 4 to 5 times longer than wide and more than seven times as long as a standard business dashboard.
Here are the Twitter accounts and Tableau Public accounts of the other eight finalists (alpha order by last name):
- Dinushki De Livera (Twitter | Tableau Public)
- Kevin Flerlage (Twitter | Tableau Public)
- Pradeep Kumar G (Twitter | Tableau Public) Top-3 Finalist
- CJ Mayes (Twitter | Tableau Public)
- Brian Moore (Twitter | Tableau Public)
- Chimdi Nwosu (Twitter | Tableau Public)
- Anne-Sophie Pereira De Sa (Twitter | Tableau Public)
- Lisa Trescott (Twitter | Tableau Public) Top-3 Finalist
2. The self-service of the classic Tableau interactive data visualizations takes a backseat to data stories
With very long visualizations, interactivity isn’t practical for user experience outside of tooltips or light filters/parameter work. Text and narration are used extensively via more charts, text boxes, annotations, or tooltips (to a lesser degree) to feed insights rather than the self-service insights used with interactivity.
As a former analyst who loves choosing their adventure through interactivity on complex data sets, it pains me to share this. Still, the spoonfed data story approach is more inclusive and can be more effective to a broader audience. The can is if the author applies the principles of a high-quality IronViz entry mentioned above.
3. Using other tools with Tableau is okay, even for a Tableau visualization contest
Every one of the Top-10 authors accentuated Tableau with a separate tool or asset not created in Tableau. The most common are images created in PowerPoint, Figma, or other devices to show icons, enhance the background or increase text options (there are very few websafe and Tableau Public-safe text options — here is a great article covering that by Ken Flerlage (Twitter | Site)).
The difference between the Top-10 authors and many authors that may not have passed qualifying is that this was done to tastefully augment the visualization and not make it a primary focus.
Bonus: All authors here are active (or are newly engaged) in the #DataFam community and use resources the community offers to upskill. Priyanka Dobhal (Twitter | Tableau Public) just shared a new article on the Tableau Journey Starter Pack, which you should check out. Although being part of the community doesn’t provide judging bonus points, the resources provided and ease of access make it quicker (and more fun) to upskill.
You worked hard but didn’t make the final or reach the Top-10.
The level of competition reflects the skill, creativity, attention to detail, and pride from such diverse points of view and different tools to extend Tableau. Many people that executed the visualization that they set out to share may not place or be recognized as a Top-10 finisher. Still, they ultimately won by augmenting skills, making time to practice techniques, and participating in this transformative experience to devote themselves to a passion project and skill showcase.
If you continue to work with Tableau, follow the Tableau community or other vibrant data visualization collective, you will be blown away by how far you can climb in a short time with additional exposure and exploration. Of course, you may still fall short Top-10 or a finalist but will realize the journey's impact by the new altitude reached and the magnificent view from the heightened platform.
