The 7 Kinds of Data Visualization People
Data visualization practitioners are a motley group, and while no two may look exactly alike, they all fall into one of 7 distinct categories.
You think everyone who does interesting data visualization is doing it with Tableau, ggplot2 or D3? Wrong. And I’m not referring to 3D pie charts. There is a long and storied tradition of bending Excel to your will to make it produce Sankey diagrams, radar plots, bullet charts and more.
Need more proof? Just check out everything Jorge Camoes does.
Then there’s the Excel of the 21st century: Tableau. You may be surprised to find out Tableau can display data in color schemes other than the default, but it can do even more than that. Tristan Guillevin made this in 20 minutes with Tableau. Live. In front of thousands of screaming Tabloonies.
But no one is ever going to win an award with Tableau or Excel. No, for awards and beauty we need to go, naturally, to Italy. From the Amalfi Coast to the sun-dappled Aosta Valley, Italy is dotted with IIB and Malofiej awards. They grow there, in fertile off-white backgrounds tended by masters like Giorgia Lupi and DensityDesign.
In contrast, the rest of the news industry seems defined in their practice by a comment from Gregor Aisch from ten thousand years ago that he has since unsuccessfully tried to walk back. Building for a mobile audience when responsive design has yet to comprehend data visualization and an audience so accustomed to print that it is apparently too distracted to be bothered to hover on a single point, their work gets more scrolly and more annotated with each issue.
And as we all know, newspapers have a lot of issues.
But the true love of static figures and text lies elsewhere. Once called “statisticians” these data scientists have brought significance tests, Bayesian priors and machine learning to industry. They also brought a disdain for spending even a penny on design. After all, who needs to read a chart when the details are right there in my thesis.
Who knows how many climate disasters could have been avoided with a tan background and ten minutes of color theory.
Non-Disclosure Agreements prevent us from showing our work publicly. It‘s great because you can refer obliquely to the immense size of the datasets, the incredible challenges posed by highly invested stakeholders, and the intersection of design and information theory. And all without ever showing any evidence that you’ve ever actually tackled any of these problems or had any impact because of that convenient NDA.
Since we can’t show any applications, the only sense of our work has to be divined from the open-source libraries released by these developers, leading the public to rightly assume that Netflix is nothing but sketchy icons, Uber is nothing but hexagonal 3D bar charts and AirBnB is filled to the brim with brightly patterned streamgraphs.
In contrast, there’s an entire subcommunity made up of actual professional people being paid actual money to create just those kinds of charts. You know them from their prolific social media, a constant drumbeat of keynotes, dive bars in exotic locales and visualization borne of some kind of twisted Venn Diagram of art, trigonometry and booze.
Just imagine how many awards they’d win if they had slightly less cluttered social calendars.
There are, technically, more than seven categories of data visualization practitioner, but some tropes were left off this list due to space constraints, blandness or fear of twitter beefs:
8. Bitter Procedurally Generated Artists
9. Sensitive Finance Visualization Artists
11. Expert Workshop Nomads
This article originally claimed Mr. Aisch’s comment occurred “ten thousand years ago”. The comment occurred in 2016 which, to be clear, was so long ago people were still optimistically using Vimeo.
This article originally referred to Tableau as “Taboloney”. “Taboloney” is not the official name of Tableau.
This article originally made a reference to Microstrategy. That was a mistake.
This article originally contained a psychoanalytical exploration of various practitioners’ relationship with Edward Tufte but it was removed when the accompanying diagram was considered to be vulgar.
This article originally provided an address for Collage College, presumably a place where academic data visualization would take place, were it to exist. Collage College turns out to have been, in reality, imaginary, and should instead have been the following: