There is no question that we are becoming more visually oriented in our approach to thinking today. You can see it in the increasing numbers of PowerPoint presentations given with the admonition that fewer words will suffice. You can see it in the increase in infographics, catchy photographs, and pictorial slogans that continue to spread across social media. And you can see the result in BI dashboards and an increasing array of visually oriented approaches to the display, digestion, and understanding of data. It is no wonder, then, that visual discovery tools should emerge as an important and rapidly growing part of BI.
Visual discovery tools are applications that typically enable non-analyst users to “play” with relationships between data items and explore an array of hidden possibilities that might yield interesting trends. They are available in some form from every major BI vendor, with a few pure play solutions leading the way. Current leaders are QlikView, Tableau, and TIBCO Spotfire, although rankings are somewhat obscured by increasing incorporation of this capacity in larger BI solutions.
The mechanics of visual discovery vary, and there are wide differences among solutions designed for simple departmental use, for small business, and for use within a data warehousing environment. Some overlap dashboards, in being able to program and create regular reports in a standard format; others are aimed more strictly at ad-hoc analysis from any data source that can be accessed (this is an area led by Tableau). But the main issue is the ability to visually discover relationships between data and dependencies without necessarily invoking sophisticated mathematical analysis. What has led to their growing popularity is a simple intuitive and interactive interface that makes them accessible to non-analysts, and provides a clear and understandable visual result. The graphic result is, moreover, automatically molded to suit the data selection.
Visual discovery tools have been available for some years, having recently received enough attention that they seemed at the height of a “hype cycle.” But these tools may continue to surprise as they bring discovery analytics — not just observation of preconceived patterns — into the mainstream. More sophisticated use will require integration with existing data stores, as is happening as analytics suites pick up on the idea. They are also beginning to move into areas of Big Data that typically seem too complex for the ordinary user to engage.
Self-service BI has always seen strong demand, although definitions and capabilities have varied widely through the past years. However, strengthening the visual presentation of information, along with interactive access, seems to play into a wide range of trends toward improved methods of visualization. Visualization creates powerful insights that can be immediately shared in today’s mobile graphics-oriented environment. While dashboards provide a passive display that is good for showing the current status of KPIs, visual discovery tools provide a way to uncover and immediately share new ideas. Even when the glow recedes from the present range of tools, this is a powerful concept for enabling new ways of thinking about data in the years to come.
Are you using more visually oriented approaches in your work? What tools do you recommend?