What is a Senior Data Visualization Engineer?

Data Visualization Engineering Skills

  • Technically, you need to rigorously hack at data visualization methods and really be able to reproduce any of the charting methods you see. This isn’t because you get merit badges for each chart type but rather because data visualization is fundamentally combinatorial and the most effective analytical applications are not a single chart but a combination of several forms of information visualization married together. You can’t make that or even plan for that if you feel like certain channels, layouts or other methods are off limits because you don’t know how to implement them.
  • Theoretically, you need to be able to understand the fundamental principles of visual display of information. You need to know why certain visual structures resonate with viewers and which are the most effective ways to encode information into graphics. It’s not enough to ape the derision of data visualization experts and proclaim “pie charts are bad” or “color is hard”, you have to actually understand color and connection and size and how they deliver information.
  • Practically, you need to think of yourself as a designer first. The real challenge with data visualization is finding out what your readers want, which only happens if you can distill the problem as expressed by their feature requests, the artifacts they’ve already produced and the actual structure of the data. This touches on interaction design, information design and graphic design. That’s a lot of design.

Do I Need a Data Visualization Engineer?

  • Large-scale applications built for multiple stakeholders across and organization who do not all share the same context or level of domain expertise. In these cases, the data visualization specialist can through interview with stakeholders and review of existing data visualization products synthesize an application that is not only as effective as the existing tools for stakeholders but is more broadly accessible.
  • Views into complex data-driven aspects of a business, especially those being supported or supplanted by machine learning, where decision makers need to trust the algorithms that are replacing their intuition. Data visualization of the performance of algorithms for the purpose of identifying anomalies and generating trust is going to be the major growth area in data visualization in the coming years.
  • Building prototypes. Along with large-scale applications, there are experimental approaches that can be better served with novel data visualization, or bespoke products meant for very specific audiences where the form can be more exotic than a traditional dashboard. Success in those cases depends on someone having the talent to be able to build complex data visualization but also have the understanding of what rules can be bent or broken in very specific use cases.
An example of a bespoke data visualization product developed at Netflix for a very particular use case where the aesthetic and function of this kind of map was well-suited, even though in normal analytical mapping applications this would be considered problematic in several ways.
  • Product Managers will typically talk about the data itself, and often fall back into the refrain “just show me the data”. They ask for download buttons so they can get a CSV or the queries that drive the view, and while ad hoc analysis will always occur, it’s these requests that are the most fertile for data visualization, because when you dive into them you find out that they perform ad hoc requests to reveal patterns that are not so easy to see with off-the-shelf tools or traditional data visualization methods. Difference charts, connected scatterplots and boxplot series have all come out of requests to “show me the data” that, when translated, really meant they wanted to see some higher order structure in what they thought could only be presented as a time series or a bar chart.
  • Data Scientists might come to you with the same questions but often come with feature requests that look like notebooks. Using ggplot2 or the equivalent, they have a preferred data visualization method that they were able to develop in an analysis of a fixed dataset that now they want to see integrated into a more broadly accessible dashboard or other internal application. This requires that you understand how to recreate what can be significantly complicated charts and, even more challenging, how to make those charts interactive and dynamic.
A chart that began its life as a static data visualization in a Jupyter notebook and was eventually deployed in an interactive form with annotations and improved styling in an internal application
  • By executives I don’t necessarily mean C-level but decision makers who look to data visualization for context and high-level insights. This could be the stereotypical “busy executive” that shows up often in data visualization manuals, but it could also be the audience for a presentation that doesn’t have the same depth of knowledge about the source material. In these cases, a data visualization engineer needs the skills and knowledge to help facilitate communication of insights and points of interest using color, annotations and other techniques drawing on visual cognition.

But Is It a Profession

--

--

--

Chief Innovation Officer at Noteable. Formerly Apple, Netflix, Stanford. Wrote D3.js in Action, Semiotic. Ex-Data Visualization Society Executive Director

Love podcasts or audiobooks? Learn on the go with our new app.

Recommended from Medium

Evaluating Benchmark with Python

Tesla and Mix Variance

Big Data and the Fashion Industry

Making every task count towards your data experience

Tax collection forecast for Minas Gerais — Brazil

3 Python Projects That Will Help Automate Your Life

Rooney is the most gender-neutral name (and other fun baby names data)

Data Carpentry for Electronic Medical Records (I)

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Elijah Meeks

Elijah Meeks

Chief Innovation Officer at Noteable. Formerly Apple, Netflix, Stanford. Wrote D3.js in Action, Semiotic. Ex-Data Visualization Society Executive Director

More from Medium

How to communicate with non-technical stakeholders

Spotify (un)wrapped

Data Preprocessing: The secret to boosting your performance

Citi Bike NYC Biking Trends