Abstract
Abstract
The purpose of this research articles is to analyze the characteristics of big data as a service (BDaaS) markets which is currently in its initial stages from adoption driving and inhibiting perspectives. This study theoretically classifies the BDaaS market into drivers and inhibitors through the PEST analysis based on previous literature. Analytic hierarchy process (AHP) matrix was used to analyze the segmentation, drivers and inhibitors, to understand the nature of the early BDaaS market. This study has proposed a new theoretical methodology for analyzing the early market of BDaaS and it is expected that this methodology may be used in the market analysis of other fields beyond the BDaaS market. Customer factors in the consumerization phenomenon and social, political, and technological factors in the PEST analysis were the important drivers of BDaaS adoption. Suppliers from the consumerization phenomenon and political factors in the PEST analysis were the most important inhibitors of BDaaS adoption. The results have important implications for policymaking and fostering other newly established BDaaS markets in emerging economies.
Keywords
Introduction
Big data as a service (BDaaS) is an outsourcing innovation that offers analysis of large and complex data sets, usually over Internet, as cloud hosted service to gain insights (Bottles, Begoli, & Worley, 2014). BDaaS is an innovation that encapsulates several big data infrastructures including storage, management, platforms, and analytical tools and techniques and provide big data-related services to customers through programable application programming interfaces (APIs) (Bottles et al., 2014). Building separate systems to analyze increasingly large amount of data is costly, time taking, and require professionals with big data management and analysis knowledge (Jagadish et al., 2014). BDaaS provides a single and flexible structure which provide common functionality of big data management and analytics to handle different types of big data (Chen & Zhang, 2014). The literature on role of BDaaS (e.g., Wamba, Akter, Edwards, Chopin, & Gnanzou, 2015) has argued for benefits for its use, including, 15–20 percent increase in return of investment (Perry, 2013), productivity, and competitiveness for private as well as public sector firms. This technology could also provide economic surplus for customers (Bechini, Marcelloni, & Segatori, 2016) and also helps decision-making which allows visibility in operations and improved performance measurement (Bertei, Marchi, & Buoncristiani, 2015). This unique feature of BDaaS has allowed the BDaaS market to grow fast (Tambe, 2014). The global BDaaS market is expected to reach US$48.6 billion by 2019, reflecting a compound annual growth rate of 23.1 percent, which considerably influence the software market (IDC, 2015). According to the National Association of Software and Services Companies (NASSCOM), BDaaS technologies in India is expected to grow eight-fold, that is, from US$2 billion current level to US$16 billion by 2025 (Nambiar, Bhardwaj, Sethi, & Vargheese, 2013; Uddin & Gupta, 2014). India is currently among top 10 BDaaS markets in the world and by 2025, India will have 32 percent share in the global market (Nambiar et al., 2013). The BDaaS market is witnessing a quick growth driven by increased demand for predictive analytics and cloud-based solutions by industries such as retail, BFSI, telecom, and healthcare (Uddin & Gupta, 2014). However, countries like India with BDaaS markets in their initial stages have faced many problems in BDaaS adoption (Dubey, Gunasekaran, & Chakrabarty, 2015). Therefore, there is need for a better understanding of BDaaS markets in their initial stages as well as their characteristics (Shin, 2015).
The purpose of this study is to analyze the characteristics of BDaaS markets in their initial stage. Externally, BDaaS markets tend to show rapid growth; however, internally BDaaS adoption experience many difficulties, such as BDaaS-related infrastructure and regulations (Jagadish et al., 2014). A clear and systematic analysis of the early BDaaS market includes both the drivers and inhibitors of BDaaS adoption. Therefore, the early BDaaS market is divided into two areas based on the research methodology of Herzberg, Mausner, and Snyderman (1959): BDaaS adoption-driving and BDaaS adoption-inhibiting. According to Lee, Chae, and Cho (2013), a market is driven by a driver and an inhibitor, whereas humans are driven by satisfaction and dissatisfaction. Therefore, examining the communalities between the dual perspectives on market drive and on human drive will help in understanding the nature of the early BDaaS market.
The segmentation of BDaaS market is conducted to classify the drivers and inhibitors of BDaaS as suppliers, customers, technology, and environments to investigate the impact of consumerization on the BDaaS market. This theoretical classification of BDaaS market is based on previous studies of BDaaS and the PEST analysis. Furthermore, for better understanding of the nature of the early BDaaS market, this study uses analytic hierarchy process (AHP) matrix to analyze the segmentation, drivers and inhibitors, classified through the PEST analysis. The results demonstrate the validity of consumerization phenomenon from the adoption-driving area perspective. Customer factors in the consumerization phenomenon as well as social, political, and technological factors in the PESG analysis were the important drivers of BDaaS adoption. From the adoption-inhibiting area perspective, suppliers from the consumerization phenomenon and political factors in the PEST analysis were the most important inhibitors of BDaaS adoption. The results have important implications for other countries wishing to facilitate BDaaS adoption.
Literature Review
The sudden rise of big data and tools and techniques to handle these data has caught the attention of many researchers (Dubey et al., 2015). However, very few research works have been carried out to systematically and clearly analyze and classify the early BDaaS market (Kwon, Lee, & Shin, 2014). Therefore, based on Herberg’s two-factor theory, this study analyze and classify the BDaaS market into two different parts: BDaaS adoption drivers and BDaaS adoption inhibitors. This study utilize various opposing characteristics of BDaaS adoption for classifying the BDaaS market in the following ways (Table 1):
The benefits of BDaaS adoption (in terms of core capacities, reduced cost, improved data quality) mentioned in the existing literature as BDaaS adoption drivers. Core competencies include the users’ convenience and ease of use. Cost reduction includes reduced IT facilities and low maintenance. Improved data quality includes completeness and consistency. The possibilities of BDaaS are many (Verma & Bhattacharyya, 2016). The risks of BDaaS adoption (in terms of data security and privacy, suppliers support, lack of availability) mentioned in the existing literature as BDaaS adoption inhibitors. The big data security and privacy, BDaaS integration with the traditional data warehouses were seen as major problems (Chen, Preston, & Swink, 2015).
Dubey et al. (2015) divided the BDaaS market into suppliers (providers of platforms and services of BDaaS ecosystem), customers (enterprises adopting BDaaS), technological issues, and environments (infrastructure, regulations, and marketplace). This market segmentation will help in understanding the factors that have significant influence on the drivers and inhibitors of BDaaS adoption. Consumerization is the emergence of new information technology/information system (IT/IS) innovations among consumers and its eventual adoption by suppliers (Lee et al., 2013).
In this phenomenon, markets is driven by consumers, rather than suppliers. Therefore, consumerization phenomenon recommends that consumer-related factors lead the markets. Consumerization is witnessed in a wide range of areas, including software as a service (SaaS), cloud computing and so on (Kabachinski, 2011). Motivated by above concept, this study will also examine whether consumerization phenomenon is valid in the early BDaaS market as well.
Definition and References of Drivers and Inhibitors of BDaaS
Analytic hierarchy process (AHP) is a structured technique that compares each factor in pairs and estimates its importance for analyzing complex decisions (Saaty, 2004). Analytic hierarchy process can solve the problems of IS outsourcing by reflecting the intuitive, rational, and irrational decisions of experts (Lee et al., 2013). Therefore, this study developed a new matric analysis method to utilize BDaaS market analysis to analyze the AHP factors from several perspectives.
Research Model and Hypotheses
BDaaS has considerable potential in emerging economies like India (Dubey et al., 2015). However, it is still in its nascent stage and BDaaS providers and customers are facing a steep learning curve (Dutta & Bose, 2015). Therefore, BDaaS adoption needs a clear understanding of the internal condition and factors of the market. In order to investigate which market factors more significantly drove or inhibit the BDaaS adoption, the early BDaaS market is segmented into customers, suppliers, technological issues, and environment.
● Adoption-driving perspective: The consumerization phenomenon in the IT market is mostly visible in cloud computing, SaaS (Kabachinski, 2011; Lee et al., 2013). This phenomenon could also be valid in BDaaS, as it is also an intangible service like cloud computing and SaaS. Therefore, this study proposes that due to the consumerization phenomenon in the BDaaS market, the customers factors will be stronger drivers compared to other market factors. This study also analyze the BDaaS market from political, economic, social, and technological perspective. In cloud and SaaS, market economic factors such as reduced costs are most influential drivers of cloud adoption (Blatman, 2009; Lee et al., 2013). Therefore, in the early BDaaS market, economic factors will also be stronger drivers of BDaaS adoption compared to social, political, and technological factors. Therefore, the following two hypotheses are proposed. H1: Customers factors will be stronger drivers of BDaaS adoption than other market factors such as suppliers, technology, and environments due to consumerization phenomenon. H2: Economic factors will be stronger drivers of BDaaS adoption than political, social, and technological factors. ● Adoption-inhibiting perspective: The privacy and security of data, and lack of standards and experts are the major concerns of the early BDaaS market (Cohen et al., 2014). Due to these problems of the early BDaaS, supplier factors will be stronger inhibitors compared to other market factors. Also, the political factors were the stronger inhibitors than others in case of cloud computing and SaaS (Blatman, 2009). Therefore, hypotheses 3 and 4 are proposed as follows: H3: Supplier factors will be stronger inhibitors of BDaaS adoption than customers, technology, and environments factors. H4: Political factors will be stronger inhibitors of BDaaS adoption than economic, social, and technological factors.
Following are the three stages of the research model:
Selection of BDaaS drivers and inhibitors (Bertei, Marchi, & Buoncristiani, 2015): In this stage, adoption driving and inhibiting factors for the early BDaaS market are selected. Selection of adoption-driving and adoption-inhibiting factors for the early BDaaS market was based on the two-theory model. In this study, the adoption drivers facilitate BDaaS adoption and adoption inhibitors hinder BDaaS adoption. Even though BDaaS drivers do not facilitate BDaaS adoption, they do not constrain it. However, even though BDaaS inhibitors do not hinder BDaaS adoption, they do not enable it. Therefore, adoption drivers only activate BDaaS adoption, while BDaaS inhibitors only hinder BDaaS adoption. Also, the drivers and inhibitors of BDaaS adoption in this study are independent and distinct from each other and selected from BDaaS adoption previous literature articles and reports (Table 1). Previous studies have been checked to determine whether the classification of factors is consistent with the definitions of drivers and inhibitors. For example, a factor is classified as a BDaaS driver if previous research and reports consider the factor as a driver of BDaaS adoption. Otherwise, it is classified as an inhibitor. Classification of drivers and inhibitors (Chen et al., 2015): Classification of adoption-driving and adoption-inhibiting factors according to BDaaS market segmentation and PEST analysis. The classification of BDaaS drivers and inhibitors according to the PEST analysis is based on Blatman (2009) methodology. Based on the definition of drivers and inhibitors, each factor is classified into market segmentation or PEST analysis. This study classified the 24 selected factors into different perspectives of BDaaS market segmentation and PEST analysis. The adoption-driving and adoption-inhibiting factors related the BDaaS market from suppliers’, customers’, technological issues, and environmental perspective and is depicted in Figure 1. In this method, the two perspectives complement each other. Also, market data from the 24 selected inhibitors and drivers were able to be used in multiple aspects. The inhibitors and drivers are classified into BDaaS market segments as follows: the supplier category denotes the BDaaS suppliers and the factors are related to their weakness and strengths. The customer category denotes the customer of BDaaS and includes the problems and benefits that customer may face while adopting BDaaS. The technological issues category denotes the issues related to technology most often suggested in BDaaS. The environments category includes the regulations, infrastructure, and conditions of the BDaaS market. According to PEST analysis, the classification of drivers and inhibitors includes political factors (refers to various environmental issues and regulations/policies related to BDaaS facing the BDaaS market). Economic factor indicates economic issues and trends, while social factors indicate those factors which influence the social and cultural aspects of the BDaaS market. Technological factor includes the technological issues and trends in BDaaS markets (Figure 2). The AHP matrix analysis based on market segmentation and PEST analysis. The AHP matrix analysis of BDaaS is composed of BDaaS market segmentation as the horizontal axis (including customers, suppliers, technological issues, and environments) and the vertical axis consists of several factors of the PEST analysis related to customers, suppliers, technological issues, and environments. The consumerization in the early BDaaS market is validated by comparing the influence of each market segment. This procedure is valid as it classifies all the factors into BDaaS market components and is applicable to the PEST analysis for examining market from a macroeconomic perspective and analyzing the internal conditions of the BDaaS market (Figures 3 and 4).




Methodology
The comparison of the total default values for the factors in each axis through a matrix analysis for testing the hypothesis is as follows:
For hypotheses 1 and 2, the default values for the BDaaS-related factors (horizontal axis) are compared for determining the most influential factors. This comparison is done by calculating and summing values reflecting the suppliers, customers, technological issues, and environmental factors. For hypotheses 3 and 4, the default values for political, economic, social, and technological factors (vertical axis) are compared using AHP for assessing the importance of each factor for testing the hypothesis.
The BDaaS factors are placed at the upper level and the PEST analysis factors are classified at the lower level. The analytical hierarchy process methodology has been used to test hypotheses 1 and 2 as the adoption drivers and inhibitors reflect the properties of BDaaS market segmentation as well as the PEST analysis factors (Figures 5 and 6).


The AHP technique is widely used for multi-criteria analyses of alternatives that can support users to examine the interaction of multiple factors in complex situations (Nielsen, 2000). The AHP technique has been used in several studies for solving problems such as ranking, resource allocation, prioritization, quality management, and benchmarking (Rose & Straub, 2001; Saaty, 1990). It is also used for assessing facilitators and inhibitors of cloud computing services (Lee et al., 2013) and e-business (Lee & Kozar, 2006). Analytic hierarchy process consists of structuring and deconstructing a problem into constitute elements and then grouping and organizing the elements into identical sets. Based on the relative importance and highest priority of each element, the hierarchical AHP model assigns numerical values to the subjective judgments. The scale developed by Saaty and Kirti (2000) has been used for assigning a number for AHP’s prioritization process. A matrix with pairwise comparisons of the attributes was used as a means for calculation. However, there is occurrence of rank reversal when a similar alternative is added to problem space due to intransitivity of decision rules and the switching phenomenon in utility functions (Shang & Sueyoshi, 1995). Analytic hierarchy process involves three stages including:
computation of the relative importance of the attributes; computation of the relative importance of all alternatives with the attribute; and priority weight determination of each alternative.
The six steps recommended for implementation of the AHP technique include the following (Lee et al., 2013):
Select the requirements to be prioritized. Set the requirements into columns and rows of the n * n AHP matrix. Perform a pairwise comparison of the requirements in the AHP matrix according to the criteria. Sum the columns. Normalize the sum of rows. Calculate the row averages.
Market segmentation and PEST analysis both have been applied to the AHP matrix analysis. The AHP matrix is composed of a vertical axis and a horizontal axis (Parthasarathy & Sharma, 2014). The vertical axis includes different factors related to customers, suppliers, environments, and technological issues arranged according to the PEST analysis property. The horizontal axis consisted BDaaS market segmentation criteria including customers, suppliers, technological issues, and environments. Comparison of influence of market component verified the validity of consumerization phenomenon in the early BDaaS market. This procedure is consistent with the objectives of this study because all the factors are classified into BDaaS market components and are also applicable to the PEST analysis. In the AHP technique, the pairwise comparison in a judgment matrix is accepted to be satisfactory if the corresponding consistency ration (CR2) is less than 0.1 (44). Consistency ratio measures how much consistent the judgments have been relative to large samples of purely random judgments (Lee & Kozar, 2006). The judgments are consistent if the consistency ration is less than or equal to zero.
A survey based on a five-point Likert scale questionnaire was constructed reflecting the AHP format for pairwise comparison of BDaaS factors. A pilot test was conducted with seven BDaaS consultants and the value of the consistency ratio (CR) was greater than 0.1 showing the insufficient consistency of the questionnaire (Lee et al., 2013). Therefore, the questionnaire was revised. The respondents selected for this survey were IT consultants who have an experience and understanding of business analytics or IT industry. The survey questionnaire was distributed to 31 IT consultants from different companies. Of these, a total of 25 surveys, whose CR < 0.1, were obtained (Saaty, 1980). According to Aczél and Saaty (1983), the geometric mean is a better factor for maintaining the reciprocal of a matrix than the arithmetic mean; therefore, the geometric mean is employed to combine the 25 responses.
Results and Discussion
AHP Results for Drivers and Inhibitors
The results provide support for hypotheses 1, 3, and 4. The results validate consumerization in the case of India’s BDaaS market. Also, supplier and political factors showed the highest default priority value and were most likely to hinder BDaaS adoption, which is inconsistent with the findings of Lee et al. (2013). However, the results do not support hypothesis 2. In the case of hypothesis 2, contrary to the findings of Lee et al. (2013), economic factors did not show the highest priority value.
The BDaaS market segment is classified in drivers and inhibitors as follows:
The suppliers represent the BDaaS suppliers and their strengths and weakness factors. The adoption-driving factors include (D1) scalability, (D2) partnership, and (D3) distribution network and adoption-inhibiting factors include (H1) lack of experience, (H2) lack of availability, and (H3) unfamiliar brand. The customer category includes the benefits and problems that customers of BDaaS face when adopting BDaaS. The customers’ adoption-driving factors include (D4) reduced costs, (D5) accessibility, and (D6) low maintenance. The adoption-inhibiting factors include (H4) distrust in security, (H5) lack of understanding, and (H6) internal resistance. The technological issues represent the technology used and issues related to BDaaS adoption. The adoption-driving factors in technological issues include open APIs, integrated databases, and multiple tenancy. The adoption-inhibiting factors of technological issues include (H7) open APIs, (H8) integrated databases, and (H9) multiple tenancy. The adoption-driving factors in the environments category include (D10) BDaaS training, (D11) enhanced marketing, and (D12) marketplace establishment. For the adoption-inhibiting perspective, (H10) lack of big data standards, (H11) lack of experts, and (H12) lack of technical standards were classified as environments factors that inhibit BDaaS adoption (Table 2).
Results for Hypothesis 1 (Not supported)
Results for Hypothesis 2
Results for Hypothesis 3
Results for Hypothesis 4
Conclusion
In this study, attempts were made to compare the results of existing research and reports to analyze the BDaaS market of India and to define the early BDaaS market based on this comparison. In the adoption-driving area, India’s BDaaS market was a government-led market system in which the government implemented different policies and scope to nurture BDaaS adoption (Verma & Bhattacharyya, 2016). Therefore, policies such as marketplace establishment and BDaaS training were suggested as important drivers to promote the adoption of BDaaS (Lee et al., 2013). However, despite the efforts of the government for the BDaaS market, this new innovation has yet to expand into the overall industry in emerging economies. Therefore, this research study focused on describing the changes in BDaaS market in India. The verification of the hypotheses depicts that the drivers of the BDaaS market in the adoption-driving area are not only related to technological and political factors but also related to social and economic conditions. This indicates that the BDaaS market has transformed from a technology-led market to a socio-economic driven marketplace. Therefore, the economic benefits, which users gain by adopting BDaaS, like collaborations and enhanced marketing, are serving as the strongest drivers of BDaaS adoption. These findings of market vitalization through socio-economic factors are consistent with the study of Ludwig, Feuerriegel, and Neumann (2015). The results in this study demonstrate that not only low maintenance costs, training, and enhanced marketing but also accessibility, integrated databases, and open APIs are also very important. However, open APIs and accessibility are much lower than other factors, which demonstrates that BDaaS market has yet to go beyond the early stage. Therefore, technological, political, and social benefits by adopting BDaaS, such as low maintenance, accessibility, scalability and so on, are serving as the strongest drivers of BDaaS adoption. In the adoption-inhibiting area, the level of services provided by suppliers in BDaaS market such as availability, security, and privacy of data are important issues (Chen, Li, & Wang, 2015). These were problems raised toward domestic BDaaS suppliers. Therefore, this research study established hypothesis 3, which proposes that factors related to suppliers would be the most inhibiting factors on BDaaS adoption. According to the analysis, along with the supplier, political factors also inhibited the adoption of BDaaS. If BDaaS market can develop collaboration and the issue of distrust in security can be solved, then during the actual adoption phase of BDaaS, technological issues, such as the lack of ability to integrate or serve individually, may emerge.
The results have important theoretical and practical implications. Theoretically, a classification method was employed based on the two-factor theory in order to analyze the BDaaS market and a new method was proposed for analyzing BDaaS market from adoption drivers and inhibitors perspectives. The results suggest that customers, social, technological, and political factors facilitate BDaaS adoption, whereas suppliers and political components hinder it. This implies that the BDaaS market is influenced by adoption drivers as well as inhibitors. This study also extends the AHP procedure by reusing the default priority values in this study’s proposed matrix. More specifically, this study proposed a new model and methodology for analyzing various factors of market and hypotheses of BDaaS market from several perspectives. The proposed methodology could be used for analyzing the market of BDaaS and other fields beyond the BDaaS market. The practical implications of this study could provide a good basis for policymaking in emerging economies with newly established BDaaS markets. More specifically, the results of this study suggest that analyzing a market from two different perspectives can be useful for emerging economies with newly established BDaaS markets. This research study evidently reveals the characteristics of the adoption driving and inhibiting area of the early BDaaS market. This study analyzes each area from various different angles by using several factors.
However, this study also has some limitations. First, very limited number of BDaaS factors were considered because of the limitation of the AHP to analyze small number of factors. Although attempts were made to include all major factors highlighted in the previous studies, there may be other important factors which could influence the adoption of BDaaS market. The respondent pool of the survey was also limited and only IT consultants who had a balanced perspective of the BDaaS market were considered as participants in this study. Finally, the BDaaS market is considered only in one country, India. Therefore, future research should consider BDaaS markets in a wider range of countries and compare them with those in other emerging economies.
The results have several implications. First, the results could provide a good basis for policymaking in emerging economies with newly established BDaaS markets. More specifically, the results suggest that analyzing a market from two perspectives could be useful for countries with newly established BDaaS markets. Second, this study reveals the characteristics of the adoption-driving and adoption-inhibiting area of the early BDaaS market and analyzes each area from various angles by using multiple factors. Third, this study could significantly contribute in fostering other early BDaaS markets since it compares the results of the study with existing reports on BDaaS market in order to describe the maturation process of an early BDaaS market.
Footnotes
Appendix: Questionnaire
| Strongly Disagree | Strongly Agree | ||||
| 1 | 2 | 3 | 4 | 5 | |
| The increased volume of data could be analyzed by scaling the number of servers. | |||||
| Collaboration between users and service providers can improve the big data services. | |||||
| Service providers could use Internet as a distribution network for boarding the big data services. | |||||
| Big data as a service requires very less investment in resources for implementation. | |||||
| Through big data as a service, users could have access to latest techniques and technologies. | |||||
| The maintenance costs for big data as a service is very less. | |||||
| Vendors should open APIs in order to collaborate with users. | |||||
| Big data as a service integrates different databases for holistic decision-making. | |||||
| Big data as a service could enable users to access data through one server. | |||||
| The development of skills required for the utilization of big data as a service includes education and training. | |||||
| The awareness about big data as a service is fuelled by Internet and social media. | |||||
| The service providers of big data as a service require an online marketplace for implementation. | |||||
| Lack of experience in developing and utilizing big data as a service is an important challenge. | |||||
| Lack of assurance for the availability of big data as a service as per the users’ need. | |||||
| Lack of information about the different service providers of big data as a service by users. | |||||
| The users doubt about the safety and privacy of data shared with service providers. | |||||
| Lack of awareness exists about the economic benefits of big data as a service. | |||||
| Decision-makers are resistive to the adoption of new innovations like big data as a service. | |||||
| Lack of accessibility of data present in different silos is an important technological issue. | |||||
| The security of data shared with vendors is an issue. | |||||
| The integration of big data services with traditional database and analytics systems could be an issue. | |||||
| The legal standards related to big data services are inadequate. | |||||
| The availability of big data services experts is insufficient in India. | |||||
| Lack of technical standards exists for big data services. |
Authors’ Biography
