May 23, 2013, 7:14 AM — A picture's worth a million data points. That's the mantra, anyway, in business analytics these days.
As the big data trend intensifies and analytics become more ingrained in corporations, the need for people who can present data in easily intelligible ways is rising. Last fall, Gartner predicted that there would be 4.4 million big data jobs by 2015, many requiring new, nontraditional skills like data visualization.
But what exactly is data visualization? Who exactly is doing this visualizing, and how is it different from creating a colorful graph or an interesting infographic? (For a deeper dive into those questions, see Dataviz: A brief how-to.)
Ironically, it's hard to get a clear picture of a data visualizer. The function is not yet well defined, and it's rare to see it as a job title in and of itself, IT career watchers say. Rather, it's a skill set that more and more companies are demanding as part of other roles, notably business intelligence and analytics jobs.
"Data visualization" as a requirement in job descriptions increased 12% over the past six months, according to Todd Nevins, co-founder of icrunchdata, a jobs board that specializes in data analytics positions. In contrast, "big data" as a requirement in job descriptions is up 63%. "Data visualization is still in its infancy but becoming more prominent as companies wrap their strategies around the extraction and usage of data," Nevins sums up.
The data that is getting visualized isn't coming from IT -- at least not so far. IT has a fairly limited role in data analysis and an even lesser role in visualization, data experts say. "IT is typically responsible for much of the dashboard and business intelligence delivery today," says Gregory Lewandowski, manager of analytics at Cisco. "But we often see IT in an order-taker capacity instead of trying to understand the end game."
A graph designed by dataviz guru Stephen Few uses stacked bars in simple colors to help viewers easily make comparisons between three sets of data. An arrow and annotation make the point of the graph clear. Click here to see a "before" image of the same graph without Few's principles applied.