A Survey of Visualizing Business Data

As businesses make the transition to digital solutions, they become overwhelmed by the volume of data they collect. The continued evolution of improved hardware propagates a cycle of collecting larger quantities of data at a lower cost. Despite the investment made to collect data, it still only accounts for a fraction of the process required for useful output. Interpretation of data is vital to unlock the potential value held within and to make the most informed decisions. Companies often employ teams of data analysts to achieve this. However, this can be very costly. In addition to the cost, it also limits the number of people who may understand or access the analysis. Employing visualisation approaches enables employees with a wide range of backgrounds to view and understand it. Opening the analysis to a wider audience encourages ideas and provokes discussion about the nature of the behaviour under investigation. This broadening of the audience highlights a unique benefit that data visualisation and visual analytics can offer.

Visualisation and visual analytics have the capability to overcome the challenges associated with large datasets and multidimensional relationships. In business, the holistic nature of "big picture" approaches is valuable, providing a complete overview of a scenario or situation. Pairing this requirement of large-scale data analytics with the capabilities of data visualisation produces fruitful and thought-provoking output.

The body of business visualisation literature is growing rapidly. During the IEEE VIS 2014 conference, a workshop entitled "Business" focused on the conversion of business data into meaningful visual insight which aids in better decision making. The workshop was so popular that a second workshop was held during the IEEE VIS 2015 conference entitled "From Data to Actionable Business Insights". In addition to these workshops, Computer Graphics and Applications published a special issue entitled "Business Intelligence Analytics".

The challenges associated with the transition to data visualisation are typically related to skill set, data scale, and ease of interpretation. Traditional visual design such as simple bar charts and line plots are often unable to accommodate the scale and complexity of data. Many off-the-shelf software packages are created to address business visualisation requirements but often require specialised training to use or are incapable of conveying proprietary data a company generates. To overcome these challenges, custom software can be developed to directly address the challenges and reduce the training necessary to use it. According to a study by Gartner, the "Visualization and Data Discovery" market segment is the fastest growing area of Business Intelligence. Inspired by our collaboration with industry, we collected literature that addresses the challenges associated with visualising business data with the aim of understanding the processes involved and maximising business output. Our contributions include:

Although the relationship between academic research and business is often complicated due to conflicts in interests and goals, this survey shows that the two are not only compatible but have a vast potential for growth in their respective fields.

This paper can act as a reference for businesses wishing to explore their own data through visualisation. Utilising both the primary and secondary classifications used in this paper, the reader can find previously published visualisation research that explores data similar to their own.