Data visualization: the power to produce an engaging, insightful experience
Context-driven and collaborative
The specific format of the visualization is chosen by Yellowfin based on algorithms that weigh the value of the visualization for a given purpose (such as a timeline versus a bar chart) but can be modified by the user. Once a graph or chart is shown, the user can click on that area and then have the option of getting more detail—for example, a year-over-year trend line. “These visualizations are context-driven, based on your previous analyses and areas of interest,” McGraw explains. If sales are showing a drop from the previous year, the user can select up to a dozen different comparisons that might have affected the outcome. “External data such as the customers’ demographics can be pulled in,” he adds, “which may account for the differences.”
A panel associated with the graph produces bullet points that provide explanations in natural language. The bullet points explain any statistically significant elements within the data. The visualized conclusions can also be shared with colleagues. Collaboration is built into the platform with a Facebook-like timeline that allows the team to share, annotate and comment in organized topics.
Yellowfin also has an unusual visualization feature, Storyboard, which integrates a presentation module with live analytics that can be automatically updated. In addition, text, video and other media can be presented on Storyboard. Designed for mobile use, Storyboard also allows for bookmarking and comments, which adds another interactive dimension.
When a visualization tool has interactive and discovery capability as well as AI-assisted analysis and presentation, the result is an engaging and insightful experience. “What we are trying to do is push this out to more people in the organization, productizing the work of data scientists and extending the value of the data model,” McGraw says.
Assessing the impact
Visualization serves a vital role not just in providing insights about the existing state but also in measuring the impact of actions taken based on those insights. “Too often we are good at creating insights, but then we either neglect to act on them or don’t follow through to measure the outcomes,” says Tapan Patel, principal product marketing manager at SAS. “Visualization plays an important role in linking the self-service aspects of preparing data and generating insights to understanding the impact from the actions taken.”
Presenting information in visualized form and explaining it is particularly important for millennials, who have grown up with visual information as a central part of their life experiences. “They are wired in a different way,” Patel says. “They not only have a greater amount of visual information at their disposal but also much more interactivity built into their personal digital lives, and they don’t want to step backward.”
SAS Visual Analytics provides a range of BI and analytics capabilities. “Besides interactive discovery and reporting, users can build predictive models while visually exploring data at the same time,” Patel explains. “SAS also differentiates itself with other analytics capabilities, including forecasting, text analysis and decision trees available through an easy-to-use visual interface.”