Data Visualization for Architects
"Data is King" is an undeniable trend in technology today, giant technology companies such as Facebook, Netflix, Google, and Amazon rely on drawing value out of the massive amount of data they have at their disposal, and this is only the beginning. A common scenario for Enterprise Architects and technology leaders today is how to display a large amount of quantitative information on a PowerPoint slide, you want the information to be easily consumed but also have the data to back up any recommendations, debates, or further inquiry.
The beginning of my foray into professional IT work was Web Design, this was during the time of AOL and Prodigy ISPs, where most folks had dial-up connections. I've always loved art, and this new medium was fun, I recall over-zealously applying Lens Flares in everything I did; ahhh the 90s. But it wasn't my first foray into digital art, CorelDraw was a huge influence on me during my childhood, the ability to create things on screen that would normally be impossible or extremely difficult on paper.
Flash forward a decade or so into the beginnings of my career as an architect, and new challenges. The common architecture language was UML and presentations were usually as exciting as watching paint dry, so I looked for new avenues of presenting data. This led me down a rabbit hole, and to my newfound interest in Data Visualization and Infographics. The motivation was "why not make this more enjoyable and easier to understand for my audience?".
However, this is also a risky proposition. I've had more success trying new techniques to disseminate information than I have failures, but one adage always remains true:
Know Your Audience
I recall a reminding experience of this truism, where I had labored for months on an IT Strategy. I was set to present to an SVP, someone I hadn't got to know very well yet, and the primary decision maker for a business unit. The deck was large, covering every aspect of my recommended approach and strategy. Forty five minutes had passed, he hadn't said a word, and in the end simply said "I never want to see a presentation like that again, keep it to three slides at most", and that was it, I was pretty devastated. I immediately bristled at the criticism and was pretty demoralized for a while, but after that I realized I had broken a cardinal rule in presenting and influencing. Ultimately I was happy I took the risk, a brief executive summary would have been better prescribed; but the deck was universally loved by the larger audience, and in the end I was reminded that taking risks can be worth it, and more importantly: Know Your Audience.
If you aren't a Statistics major, and I am not, you may not have heard of some of great influencers in the domain of data visualization. I was fortunate enough to see a seminar by Edward Tufte, professor emeritus of political science, statistics, and computer science at Yale University. It was an inspiring seminar, and Mr. Tufte has a wealth of experience. He was sort of a diva arriving and departing in a town car even though he only lived an hour away, and some of his outlooks were a bit dated; I doubt he could have foreseen the advent of Big Data. Some of his advice was haphazardly delivered from a position of absolute authority, which most architects do not have. His knowledge, principles, standards, and experience however are monumental; what I found most enjoyable was his way to relate basic human logic to visualization, and the historical reasons behind today's modern techniques.
On the opposite spectrum, David McCandless is a famous infographics, computer science veteran, and focuses on publication; with visually appealing simplistic approaches. He's done quite a bit of work for BBC, CNN, and various news outlets. What I've learned over the years is, there is no surefire way, but there are surely many intelligent and creative people thinking of why and how to present data. His book "Information is Beautiful" is a fantastic approach to applying a wealth of data to simple infographics.
There's still so much more to learn. Below are some links to various tools, personalities, and sites that have helped me become a better presenter and ultimately a better architect. A portion of being a successful architect is influence, and when you are trying to influence a decision maker towards an objectively better outcome or educate an organization, any leg-up is beneficial. Data Visualization is one of those differentiators.
Microsoft Visio, is the defacto tool of technology architects worldwide, when not in an organization using a larger EA platform. Although I don't call out Visio resources here directly, I would highly recommend learning the Data functions of Visio Pro for the distinct advantage of linking spreadsheet data directly to a model, I've found this particularly useful when maintaining large complex diagrams and roadmaps.
- David Mcandless: https://www.ted.com/talks/david_mccandless_the_beauty_of_data_visualization#t-1074086
- Ed Tufte: https://www.youtube.com/watch?v=rHUDJ8RyseQ & his website: https://www.edwardtufte.com/tufte/
- Additionally, this blog post is a very good summary of his seminars: http://www.duskbeforethedawn.net/2015/05/notes-and-thoughts-on-edward-tuftes-one-day-seminar/
General Data Visualization Sites
- Information Aesthetics, examples of data visualization : http://infosthetics.com/
- Visual Complexity, various examples by subject/genre: http://www.visualcomplexity.com/vc/
- FlowingData, website/blog about DV: http://flowingdata.com/
- WikiViz, wiki and tools list for DV: http://www.wikiviz.org/index.php/Tools
- D3.JS: https://d3js.org/
- Gephi, Open Source Data Visualization Tool: https://gephi.org/
- Circos complex data visualization tool, used for DNA and gene data visualizations: http://circos.ca/
- Cubism.js, time series animated chart: http://square.github.io/cubism/
- R, stats programming language that feeds into other tools: https://www.r-project.org/about.html
- Sigma.js, another graph drawing JS library: http://sigmajs.org/
- Visual Literacy, What visualization should you use?: http://www.visual-literacy.org/periodic_table/periodic_table.html
One of my favorite newer discoveries. CloudCraft is significantly tailored for AWS, and more importantly lets you link up real data from AWS directly to your models. https://cloudcraft.co/