1. Introduction

Over the last decade, there has been a tremendous increase in the amount of digital data accrued from various sources. Database systems allow for a mass-collection of data which can be used for problem detection and prediction, but also for gaining an increased overall understanding of current business models by identifying important influencing factors. Consequently, we can observe an increased interest in using these large structured and unstructured data sets in order to enhance informed and analytical decision-making.

This trend has also influenced accounting professionals. Handling large amounts of data is not a new requirement as at a fundamental level the profession has to summarize, structure and prepare data for various decision-making purposes. However, the concept of Big Data is increasingly not only understood as the handling of data high in volume, but also high in variety and velocity. This is because not only is company internal data used for analysis (mostly structured data) but also external sources such as websites, texts, videos, the Internet of Things, RFID, sensors and other items/sources (mostly semi-structured and unstructured data), which are required to generate a more comprehensive picture. Nonetheless, we speak of Big Data, as soon as one of the above-mentioned criteria is met.

In this context, the discipline of information visualization (InfoVis) has gained increasing attention because its goal is to identify and create distinct images by emphasizing particular task or data characteristics to facilitate understanding and comprehension. Visualization per se is particularly helpful in this regard and especially for management accounting as the objective is to inform internal and external stakeholders about the past, the current and the future state of the company. By means of visualization, trends, correlations and irregularities can be localized in a more efficient and effective way. This is especially true if the data sets are increasing in size and complexity. To do so, reporting in various forms (e.g. internal and external) has been institutionalized in accounting and the use of traditional visualizations (for instance pie, column or bar charts) is already common practice.

However, with the change in the data structure brought about by Big Data use, researchers and domain experts argue that the way various stakeholders are being informed needs to be altered. Including not only structured but also semi-structured and unstructured data sets, led to the development of newer forms of visualizations better suited for the new requirements. Recently, these new and interactive visualization options are generated on a regular basis, nonetheless, studies indicate that there is a considerable gap between what is possible and recommended from an expert’s perspective to what is actually employed by practitioners. Although the value of the new visualization forms in the context of Big Data is recognized, it seems to be rather a preferred analytics topic on the wish list for future implementation than something widely used. Consequently, users remain to rely on simplistic and traditional visualizations for complex problems, which erroneously simplifies the related information of data sets. Unfortunately, this can cause interesting relationships to remain hidden, decision-making options to be reduced, and users to be encouraged to rely on biases and heuristics during the decision-making process.

The purpose of this paper is twofold: first, we investigate the status quo of reporting and analyze whether the incorporation of semi-structured and unstructured data sets have already taken place and whether this change has caused modifications in current reporting practice with regard to visualization use. With respect to modifications made, we are not only interested in the adoption of newer visualization types but also in the adoption of interaction techniques, as the latter is also a key component for sense-making in early Big Data visual analytics processes. Second, we focus on reasons regarding the previously mentioned slow adaption. We contribute to the ongoing discussion by being the first to investigate in detail why this obvious gap between experts and practitioners exists. To do so, this paper will systematically distinguish between human-related and technology-related barriers from a finance and accounting-related perspective. While for human-related barriers, we concentrate on the perception and experience with new visualization forms, we investigate functionalities and characteristics of tools and data sets for technology-related ones. Gaining an understanding on barriers may help in overcoming them and enhance the use of interactive visualizations for Big Data, which is highly relevant for sense-making in a data driven environment. Further, this is one of the few studies that concentrate on actual users and their needs rather than introducing and promoting new visualization options, which is current practice in the field of InfoVis.

Empirical evidence is collected using a quantitative questionnaire, which is distributed to accounting professionals in Austria. In total, 145 evaluable responses from a broad variety of business sectors could be obtained and were used for analysis. In order to distinguish the visualization types under investigation look at the following tables and figures. While Figure 1 represents visualizations encountered in everyday life (which are also the main types used by accounting professionals in the past and in the further course of this paper defined as type I visualizations), Figure 2 represents new and more complex visualization types to be used in an interactive manner especially designed to handle large structured and unstructured data sets which are henceforth called type II visualizations.