Katherine Fraser, manager of data analytics and visualization at Volvo Group, speaks to Innovation Enterprise about the challenges of creating viable data visualizations, the tools available to designers and visualization specialists, and the role of visualizations in helping leaders to make smarter business decisions.
Innovation Enterprise: Does data visualization in 2018 require specialist training and knowledge, or can it be made accessible to relatively nontechnical users?
Katherine Fraser: This depends on whether you’re talking about consuming data visualizations or creating data visualizations. Being able to tell a story with data requires a diverse set of skills including the ability to gather, cleanse, and relate data (this typically means coding, such as SQL and R), the graphic design eye to be able to choose the chart type, layout and colors that most effectively convey the information, and the creativity to come up with the best way to reveal insights.
While there are tools today that are targeted toward business users and allow a nontechnical person to easily create a graphic, employing a trained visualization person will yield better results. And better data visualizations mean they are easily accessible for anyone to see and understand.
IE: What data visualization tools are available? Are there compatibility issues across different toolsets?
KF: You can use many different tools to create pictures with data. In the hands of a skilled visualization person, even general office productivity applications like Excel and Visio can be used to make nice graphics.
Today, there many different products targeted more specifically at creating data visualizations. Tableau has always been focused on making beautiful charts. QlikView is another common choice for creating dashboards made up of different chart types. I’m a big fan of Microsoft Power BI because it’s powerful, yet easy to use and supports custom visuals. Any BI tool — SAP Business Objects and Lumira, GoodData, MicroStrategy, Pentaho — will allow you to visualize the data you bring in.
Compatibility isn’t really an issue since the real work is done prior to visualization, gathering and shaping the data. And once the data is prepared, you can import it into any tool to present it graphically.
IE: What role can automated data visualization tools play in helping to sustain a high level of business intelligence and employee productivity?
KF: Automation could mean a couple of different things in the context of data visualization tools. In a more traditional sense, an automated tool might send alerts when predefined thresholds are reached or might email graphical reports to a list of consumers.
In another context, a visualization tool might try to choose and display charts automatically based on an intelligent reading of the data. AI can be used to interpret the data and guess at what kinds of analysis are useful in deciphering the information.
Both types of automation can be advantageous but depend on the quality of the tool. Too many false positives and users will start to ignore automated alerts. And misreading the data produces useless automated analytics. I think we’ll see improvements on automation in the future as machine- learning algorithms get better at finding insights in datasets.
To find out more about the latest Data Visualization trends for the coming year, download our free ebook: DATAx Guide to Data Visualization in 2019