Contemporary surveys of IT decision-makers have consistently ranked business intelligence (BI) among the top items on their agendas. Interestingly, however, data visualization has taken a backseat in terms of mind share and demand within the BI space. In contrast, data warehousing, data mining and analytics are all viewed as fundamental facets of the growing BI craze.
To test this hypothesis, I performed a simple analysis on the relative popularities of key constituents of BI, starting with the ubiquitous Google search. First, in a search for "business intelligence," Google returns 150 million results. A second search for the combination of "business intelligence" and "data warehousing" returned 42.9 million results. If you then replace "data warehousing" with "data mining," the results are further narrowed to 12.6 million. If you substitute with "analytics," it returns 12.1 million. Finally, if you try the combination of "business intelligence" and "data visualization," the search only returns 1.2 million results. Thus, within the bounds of the commonly used Google search, the hypothesis is supported: data warehousing, data mining and analytics are more heavily discussed topics in the BI space than data visualization.
Visualize the results shown in Figure 1.
Despite its low Web profile, data visualization is gaining recognition among C-level executives and is key to the success of any BI initiative aimed at the general business audience.
Let us ask the following three questions. How many end users would consider the effort and complexity behind creating a data warehouse infrastructure? Similarly, how many end users would take into consideration the complication of queries that generate their daily reports or the sophisticated analytics behind them? In contrast, how many end users are concerned with how the data is presented to them and with how easily the reports can be comprehended?
In most cases, I would bet that business users are less concerned with the behind-the-scenes coordination that they are not directly involved with. Very few decision-makers will be interested in seeing the raw data. Of main concern to business users is the end product they are visually presented with and the responsiveness and accuracy of the reports. Not to imply that all the work and complexity behind data warehousing, data mining and analytics are any less important for the success of an organization's BI initiative, but the end user's perception and adoption is driven by the end product that gets delivered. The end-product delivery happens at the data visualization layer.
In many cases, the delivery of BI to end users happens through the reporting layer. The reporting interface has stepped up and developed user-friendly features, including the idea of ad hoc drag-and-drop reporting. The data visualization has been long embedded within reports through the charting options within a reporting framework. Ironically, because it is perceived as being a small slice of many reporting features, data visualization never receives due respect as a key pillar to the success of any BI delivery platform.
The key development that has changed the perceived role of data visualization is the growing popularity of dashboards. Dashboards are the new face of BI, and dashboard vendors have focused a great deal on data visualization to build a more effective and pleasing user experience. Furthermore, traditional reporting vendors have enhanced their existing reporting frameworks with added versatility to the embedded charting options, whereas pure-play dashboard vendors have often developed a user experience completely immersed in interactivity and rich visual animation, much different than the traditional reporting aesthetics. It does not take much speculation to guess who wins the end users' vote.
In either case, the common winner is the field of data visualization. It is rapidly getting recognized for the value it brings for delivering BI to the business users. In the absence of an effective data visualization layer, most of the BI investment would stay locked within a gated community of a few trained and technically savvy BI consultants.
While data visualization gets vindicated as the key success element for BI, much to the dismay of purists and researchers in the field, it is often not the best and the latest research that wins the audience.
The new genre of popular BI data visualization is an interesting concoction of the most basic data visualization features with a mix of animation and interactivity, blended with a simulation of speedometers and thermometers. All it takes is a study of Edward Tufte's treatise "Beautiful Evidence or The Visual Display of Quantitative Information" and comparing these landmark studies in data visualization within the context of real-life business adoption. In many instances, the comparison is similar to the parallels between Shakespearian poetry and a contemporary business article.
It is the same explanation that helps to clarify the divergence in both instances - Shakespeare's literature from the language of the common masses, as the advanced data visualization from the BI's masses. In the course of conducting business, people do not have the inclination, time or interest to appreciate the depth and intricacies behind some of the hard-driving results. Instead, they want simplicity, while they continue on with their regular daily roles and responsibilities.
Consider the following examples that represent the applications of BI to the masses. In a call center, the shift managers need to monitor relative performance of the operators, average call handle time, calls in queue, abandoned calls, etc. In a health care facility, the staff needs to monitor customer satisfaction, individual performance, emergency responsiveness, etc. In a sales organization, sales representatives and their managers need to monitor their sales pipeline, actual versus forecast, top product sales, etc. The common thread across all these scenarios is that the audience needs quick access to relevant information without being distracted by any intricacies of presentation. The ability to quickly obtain clear, intelligible data allows business users to stay well informed and prepared for the job at hand. The key again is simplicity, which data visualization facilitates.
