Being at a Tableau conference is a seriously heady experience. It's like attending a wedding for an insanely popular couple who invited 17,000 of their close friends. The energy, the music, the food, the quick, often drive-by conversations with people you'd love to talk to longer. I always leave them exhilarated and inspired to do better - better work, better community engagement, better nourishment of friendships I've already made. And this year those feelings were particularly acute since lately I haven't been as engaged with this community as I would like to be. Switching to a new job has consumed far more of my mental energy than I expected it to. Perhaps it was naive of me to expect that not to happen, but that naïveté was likely the result of having been at my previous employer for a VERY long time (21 years). I knew where all of the bodies were and I knew who put them there. I knew which ones needed exhuming and which ones were better left alone. Having been there for two decades, I was at the epicenter of tribal knowledge and benefited from the insider shorthand that went along with that. So I could essentially coast and still produce solid work, which then left me with sufficient time to engage regularly with the community.
That all changed in January of this year when I started at Pluralsight and bellied up to the fire hose of new data, people and processes at a rapidly growing company.
I am beyond grateful to be there and for the opportunity to work with and learn from such a talented and generous team, but I do miss my regular participation in the Tableau community and this year's conference only amplified that.
It also helped me crystalize something I've been thinking about for a while, namely how I connect with people, or fail to, as the case may be. And in particular, how that carries over into my data work. As an introvert, I'm not particularly adept in social situations, especially large ones. You'll often find me standing about, a lumbering giant trying to think of a good response to an aborted conversation I had with someone half an hour earlier. Or figuring out what to say to the person about to enter my orbit before the gravitational pull of a more interesting person steals them from me. I think I struggle in these large group settings simply because it takes time for me to figure out which questions to ask people, or which aspects of myself to share to establish a meaningful connection, and it's just very easy to get overwhelmed by the pace of conversation or to feel like you're competing with others who appear to have access to some special superfood that provides them with extra doses of gregariousness and charm.
So what does all of this have to do with my data work?
Well, for me, choosing to get involved in a data initiative can sometimes feel like walking into a very large room where there are so many other people already having amazing conversations. There's an immediate and strong desire to make a connection and get myself included in those conversations. And in the data world, unlike in the physical world, I can sometimes find that elusive superfood and produce visualizations that are eye-catching and pull me in from the periphery. I get my little moments to strut.
But I think this desire to be noticed and included can sometimes distract me from actually communicating anything meaningful. I get so caught up in trying to produce a striking or novel design, or engaging with as large of an audience as possible (in the form of likes and retweets) that I can manage to say nothing at all. Meaning, I either forego contributing altogether, or just come out with a shiny and hollow bauble that fails to resonate with anyone.
This isn't to say that every visualization has to actually say anything. Frivolity and experimental play are critical to growth and to nurturing the excitement that drew us to data visualization in the first place. It's why I am so in awe of what people like Neil Richards and Ken Flerlage do. They clearly enjoy their hobby and seek out opportunities to improve their skills via intentional play, something I want to get better at. However, when it is time to say something, I need to combat the impulse to be noticed and not allow it to guide me. I need to be okay with taking the necessary time to get to know the topic and find the fact or anecdote that means something to me, that moves me, and then let that be the spark that informs my approach to the visualization. This approach may not result in a visualization that resonates with a large audience, but if it genuinely conveys my reaction to the data, then I've got a much better chance of making a connection with at least one person in the community who may respond to the data the same way I did. An audience of one doesn't have to be a failure; it can, in fact, be the goal.
What does this look like in practice? A project I contributed to Viz for Social Good earlier this year is a good example.
It was a project on homelessness in England, partnering with an activist named Papa Baiden who has dedicated his life to raising awareness of the homeless and by doing so influencing policy change. As I was reading his website, I noticed that he talked a lot about changing the way people view the homeless as a critical part of this project. And I realized pretty quickly that that referred to me as well. While I didn't harbor any negative views about them, I knew nothing about the homeless in England. They were just an abstract headline to me. Calling people "abstract" sounds cold and mildly sociopathic, but in our data-rich culture, I think a certain degree of numbness becomes a survival tactic. You can't care about everything. At least I can't. While I like to think that I'm a fairly empathetic person, I know I've walked by suffering more times than I care to admit.
My initial impulse was to come up with some design that simply got the data noticed. Something visually riveting that would, hopefully, get people to at least pay the topic a glance of interest. But nothing about the data drew me in, and while I flew through several design sketches in my mind, none of them stuck around long enough to get birthed. While the idea of coming up with something eye-catching certainly appealed to me, that appeal wasn't enough to fuel the work it would take to actually put it together. So at this stage, I wasn't sure I was going to contribute. I didn't feel like I had anything to say, due to my ignorance about the topic, and the idea of pursuing a design that was as much about itself as it was about the data didn't excite me. If I wanted to make an impact, to connect with not just the Tableau community but the larger community that Papa Baiden planned to reach, I was going to have to find an emotional hook, something that would invest me in the topic.
When I turned to Google for some additional context and inspiration, I initially perused sites that provided more data and analysis about the homeless situation in England. Good information, but I still felt that I was just marinating longer in an abstract subject. And then I popped over to the Image results, since I have often gotten design cues from photos. In so doing, I saw a lot of faces of people who were currently homeless in England. And most of those photos linked out to sites (click the photos below to see some examples) that provided their stories - how they became homeless and what their lives were like now. I read one. Then another. And before I knew it, I had spent a few hours reading about dozens of people I had never met but whose stories were encoded in the data I was trying to visualize.
Just like a landscape photograph can be made more accessible by adding a human into it (something I try to do as often as patience and serendipity allow), because it allows the viewer of the photograph to imagine themselves in that scene, I knew that this data set would be far more accessible to my eventual audience if I could emphasize the human element inside of it. It is what ultimately engaged me, and so that is what I wanted to focus on.
I got the idea of assembling a human face out of a ton of data points, to really drive home that theme. Thanks to a blog post by Curtis Harris that I recalled reading many months ago, I was able to use some free software to turn a photo of a homeless man into a data set of over 1 million rows. Once I had imported that data into Tableau, it was just a matter of some scatterplot and bin tweaking to get the face to appear the way I wanted it to.
My initial grand vision for this was to actually have a story per row of data - so a million different stories that someone could click into from my visualization. I thought that that would have made for an awesome interactive tapestry. But I had to dial that ambition way back due to time and so instead added a column to the data set and randomly assigned roughly 1,000 first names (half female names and half male names) to many of the rows. This allowed me to place those names in the tool-tip for the sheet containing the man's face, so that those names would appear as someone hovered their mouse over the image. But since the stories are what hooked me and provided the inspiration for the design, I wanted to make sure they made their way into the visualization. So I selected six of those stories and added them below the man's face, and then used the tool-tips to provide information about each individual along with a link to the website where their full story could be read.
The rest of the visualization contains the data, the actual information that we want people to digest and be informed by. The charts are very functional and simple - three line charts, a bar chart and a map - and are laid out in a way that emphasizes the data and de-emphasizes the non-data. But because that data didn't grab me before I took the time to get familiar with the people they described, I felt that the data should be secondary to the human stories and therefore in the bottom half of the visualization. I know there are people out there who already care deeply about the homeless, or know some of them personally, or approach this topic from an intellectual perspective, all of whom would likely not need the large face of a homeless man or the individual stories to get them engaged with the data set. But I was none of those people, and I was creating this visualization for people like me who had no entry point into this topic. For that reason, the human connection had to supersede the numbers.
 |
| Click the image to go to the interactive version |
Later that evening, after our talk, I was sitting in a nice large chair in the lobby of the hotel, having left the big party because it was just too loud to even attempt a conversation with anyone. A lady walked up to me and said that she had been in our session and had enjoyed our discussion and that my visualization had really resonated with her. She had always considered getting involved in projects that used data for social causes, but had never known how to begin or what that might look like. Now she did.
Six months after producing my viz, I connected with an audience of one. Mission accomplished.
Thanks for reading.
View comments