Previous research and definition of key concepts
Even though the focus of this investigation is not the exaplanation of the business intelligence market as a whole, I believe it is worth providing a brief introduction and explanation to the argument, due to the unfamiliarity that many have with some terminologies or processes. In this way, I have the opportunity to clarify some parts and guide the reader toward the focus of the research.
I deliberately decided to divide this section in two distinctive parts, each of them dedicated to the review of a specific area. In the first part I introduce the definition of small and medium sized enterprises (SMEs). The second part is about the Business Intelligence domain. Here, several sub-sections are presente: the history of Business Intelligence; a short explanation of Business Intelligence systems; a comparison between and cloud SaaS vs traditional BI solutions.
Small and medium sized enterprises
A unique definition of small and medium sized enterprise has not been decided yet (Carter and Jones-Evans, 2006) and, for the purpose of this study, a company is considered a SME if it fulfils the following requirements:
• Up to 500 employees and $25 M in annual revenue in the Unites States (Carter and Jones-Evans, 2006).
• Less than 250 workers; a maximum annual turnover of €50 million or €43 million in the balance-sheet, for european enteprises (Carter and Jones-Evans, 2006).
• For asian companies there is not an official definition and it varies greatly from country to country. For instance, chinese companies with 2000 employees can still be considered medium businesses, whereas in Lao, a company with more than 100 employees is considered a big company (Harvie, 2004; Xiangfeng, 2007).
Since this study encompasses businesses belonging to different countries, I will use all the three the above-mentioned definitions.
Evolution of business intelligence
Business Intelligence is a term that may seem unknown to some, but it has roots long back in the past. It is commonly agreed the term Intelligence has been coined the first time by IBM analyst H.P. Luhn. He defined it as “the ability to apprehend the interrelationships of presented facts in such a way as to guide action towards a desired goal” (Luhn, 1958). Therefore, the first decision-support systems were born and further developed in the following years, becoming important IT solutions for supporting the decision-making process. In 1989, Howard Dresner proposed the widely accepted definition of Business Intelligence, still used today; ”concepts and methods to improve business decision making by using fact-based support systems” (Watson and Wixon, 2007).
Definition of business intelligence
After a detailed analysis of the literature, there exists so many and different definitions of BI that is fundamental to pick one and stick with it in the whole paper. Hence, with the term Business Intelligence I refer to the “processes, techniques or tools that support the making of faster and better decisions” (Pirttimäki and Hannula, 2003).
Taking as a reference the number of articles being published in the last 24 months, the hype around Business Intelligence has increased significantly, indicating that BI systems represent an important component of a modern enterprise’s information infrastructure, as they contribute to its success and competitiveness (Davenport et al., 2010). Moreover, the possibility to dig deeper into the consumer behaviors by analyzing multiple sources of data has become a central concept and activity within an organization (Nyblom et al., 2012). This study is not aimed at describing the business intelligence software in details, but clarifying some simple concepts related to the main functionalities is crucial to grasp the further sections of the paper. Historically, it was essential to not confuse the terms business analytics and Business Intelligence. They were not interchangeable and the best explanation of the differences has been given by Vesset and McDonough (2007) through the use of a self-explanatory image (Figure 1).
As the figure shows, the BI tools/software represented only a small set of a broader market segment called business analytics. In particular, we refer to the Business Intelligence application as the category involved in querying, reporting and analysis of data through the use of complex techniques.
Things started to change in the last years when the interest around the business intelligence market rose significantly, with a substantial number of new companies entered into this sector. This rapid increase might have caused the misuse of the term Business Intelligence, which has been referred wrongly to other types of software (Søilen and Hasslinger, 2012). Moreover, it is possible to find companies, that claim to offer Business Intelligence products, whose characteristics do not belong to any specific category defined by Vesse and Mcdonough (2007). As a consequence, the business intelligence market has been evolving in a more business homogeneous environment (Søilen and Hasslinger, 2012) where the boundaries defined by Vesse and Mcdonough (Figure 1) have been crossed many times. Given the previous considerations and for the purpose of this study, I consider appropriate to provide a definition of the BI systems: “BI systems combine data gathering, data storage, and knowledge management with analytical tools to present complex internal and competitive information to planners and decision makers” (Negash, 2004, pp. 178).
BI systems’ architecture
One of the main benefits offered by business intelligence solutions is the representation of structured and unstructured data in an easy and understandable format. Yet, the use of graphic and visualization tools to “make meaning” of data is only the final result of multiple processes taking place within a complex architecture. Details are not necessary for this research, but I want to provide a schematic overview of these processes.
Cloud SaaS vs traditional BI solutions: part one
From the late 1990s, many companies started using BI systems for two reasons: keep their knowledge stored in a single place and leverage it to gain better understanding of the business environment. Historically, the implementation of a BI software has been mostly a privilege of the big companies, who could afford the high costs that characterize these IT systems (Scholz et al., 2010; Olszak and Ziemba, 2012; Wong, 2005). As a matter of fact, the traditional BI systems were built on-premise, physically installed and run in the building of the company using the software. Therefore, on-premise BI systems were tailored to the specific needs of the company, and this approach often had costs not sustainable for a small or medium sized business. For the sake of clarification, in this paper I will use interchangeably the terms on-premise and traditional BI systems.
In the last years, a new way of utilizing software has been taking place: “Cloud computing is a model for enabling convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction” (Feenstra et al., pp. 3).
In general, cloud computing can be classified in three different layers: Software-as-a-service (SaaS), Platform-as-a-service (Paas), and Infrastructure-as-a-service (Iaas) (Zhang et al., 2010, Feenstra et al.), which represent the different delivery models for the cloud services (Östling and Fredriksson, 2012) (Figure 3).
The theories used for the analysis of the empirical findings are structured around two themes which describe the key adoption factors for Business Intelligence software in an organization. Since SaaS Business Intelligence in SMEs is a rather new field of research, the sources of information I used are quite disparate. First and foremost, as suggested by Bryman and Bell (2011), I leveraged online databases and scientific journals such as ABI inform, Google Scholar, Journal of Computing, Journal of Intelligence Studies in Business and Journal of Accounting, to name just a few. These sources offered the right overview of the Business intelligence market. Another step, whose importance has been remarked by previous scholars (Bryman and Bell, 2011), was the careful choice of keywords. The most used in my research were: cloud, cloud software, Business Intelligence, Business Intelligence SME, SaaS Business Intelligence SME, cloud Business Intelligence, success and choice factors BI (SMEs), implementation factors SaaS BI, selection criteria IT software, evaluation criteria BI in SMEs. Once an article was found, close attention has been paid to the references, and this snowball approach allowed me to find other valuable sources unknown at that point. Articles focused on the success factors proved to be particularly useful, such as Adamala and Cidrin (2011) and Olszak and Ziemba (2012), while Søilen’s studies offered a constant guide throughout the project. Regarding the structure and language of my study, Vodapalli (2009) and Yeoh & Koronios (2010) have been a great source of inspirations.
Table of contents :
1.2 Problem area
1.3 Purpose and research question
1.4 Chapters alignment
2 Previous research and definition of key concepts
2.1 Small and Medium sized Enterprises
2.2 Business Intelligence overview
2.2.1 Evolution of Business Intelligence
2.2.2 Definition of Business Intelligence
2.2.3 BI systems architecture
2.2.4 Cloud SaaS vs. traditional BI solutions: part one
2.2.5 Cloud SaaS vs. traditional BI solutions: part two
3 Literature review and research design
3.1 Key adoption factors: definition
3.2 Chapter’s structure
3.3 Key adoption factors in Business Intelligence
4 Research Methodology
4.1 Research purpose
4.2 Research design
4.3 Research approach
4.4 Research strategy
4.5 Literary sources
4.6 Empirical material
4.7 Sample and limitations
5 Results and analysis
5.1 Stage one: categorization of key adoption factors and discussion
5.1.1 Key factors from the qualitative interviews
5.2 Stage two: findings and discussion
6.1 Practical implications
6.2 Theoretical implications
6.3 Shortcomings of the study
6.4 Suggestions for further research