INTRODUCTION

Traditionally, procurement value lies in cost reduction, through tactics and competitive contracting (González-Benito, 2007; van ^{Weele & van Raaji, 2014}) and a growing number of studies have been demonstrating the need for moving forward to a clear value proposition strategy (^{Hong & Kwong, 2012}; Chick & Handfield, 2015; ^{Lilien, 2016}). Procurement has been shifting its tactical role to a core element for strategy and competitiveness of organizations (Palraj et al., 2006; ^{Tassabehji & Moorhouse, 2008}). Most operational activities are already being rapidly digitized (^{Arnold et al., 2005}; Hong & Kwon, 2012; ^{Nicoletti, 2013}) to support or substitute procurement operational tasks and it leaves free time for the development of more strategic action (^{Presutti Jr, 2003}; ^{Bendoly & Schoenherr, 2005}; ^{Wang et al., 2016}). In this sense, technology in procurement activities is a way to connect and shift relationships (^{Gebauer & Segev, 2001}, ^{Cuganesan & Lee, 2006}), achieve cost efficiency by optimizing process (Arnold et al., 2005; Hong & Kwong, 2012), exploit the procurement roles (e.g. ^{Gadde & Håkansson, 1994}; Tassabehji & Moorhouse, 2008) and enhance skills and capabilities (^{Carr & Smeltzer, 2000}; GonzálezBenito, 2007). It opens the doors to new organizational configurations and analysis (^{Nollet & Beaulieu, 2010}) by requiring procurement and organizational integration and alignment (Ellram & Carr, 1984; Carr & Pearson, 2002; ^{Pohl & Förstl, 2011}; ^{Ateş et al., 2018}), through the identification and translation of business competitive priorities to clear procurement actions to fit business strategy continuously (^{Narasimhan & Das, 2001}, ^{Gonzalez-Benito, 2007}).

However, the size, variety, and speed of data that impact procurement decision-making are expanding fast, and, the more strategical procurement became higher the demand for taking real-time assertive decisions. The evaluation of big and complex data in deep offers unprecedented opportunities for advancing in procurement value-adding, lower costs (^{Chae et al., 2014}; ^{Wang et al, 2016}) and avoid fraud (^{Ramamoorti & Curtis, 2003}; ^{Westerski et al, 2015}). On the other side, practitioners need to discern and act fast (Chick & Handfield, 2004), avoid premature design, share, and take improper conclusions (^{Van Knippenberg et al., 2015}; Bendoly, 2016). It requires investments in technology, organizational process change (^{Barbosa et al., 2018}) as well as strategies and specific capabilities to handle and transform such huge amount of information into clear, reliable and valuable material to leverage resources (^{Tippins & Sohi, 2003}; ^{Kiron & Shockley, 2011}; Carrillo, 2017).

Despite the strong interest of big data in supply chain and procurement strategies (e.g. ^{Chen et al, 2012}; ^{Waller & Fawcet, 2013}; ^{Sanders, 2014}; Li et al., 2016; ^{Kache & Seuring, 2017}; ^{Gunasekaran et al., 2017}; ^{Roßmann et al., 2018}), theoretical and empirical evidence of how to include big data strategies into procurement managers decision- making process still lacks (^{Mogre et al., 2017}). Based on that, this study intends to verify: 1. To what extent do alignment, knowledge, and coordination of big data strategies impact its use and efficiency in the procurement environment? 2. Do current procurement executives’ behaviors (intention and reaction of individuals) impact the deployment of data analytics and consequently the overall superior results?

To address these questions, we structure our paper as follows: First, we review some work exploring the integration between corporate, procurement and big data analytics strategy, followed by the role of procurers’ behavior (intentions, actions, and reactions) that should influence the deployment of data analytics. Then, we present the methodology and data analysis. We conclude by discussing the implications of our findings and propose possible future studies for advancing this discussion.

INTEGRATION OF BUSINESS, PROCUREMENT, AND BIG DATA ANALYTICS STRATEGIES

“In the coming years, most intelligent organizations will need to blend technology-enabled insights with a sophisticated understanding of human judgment, reasoning, and choice. Those that do it successfully will have an advantage over their rivals.” (^{Schoemaker & Tedlock, 2017}, p. 27)

It is well discussed in the literature that, to succeed, Procurement must be aligned to corporate strategy (e.g. ^{Watts at al., 1995}; ^{Narasimhan & Carter, 1998}; ^{Knudsen, 2003}; ^{Gonzalez-Benito, 2007}; ^{Baier et al, 2008}). As Ellram & Carr (1984) state “Purchasing must assume a proactive role in working with other functions to formulate and implement competitive strategy, to minimize barriers between purchasing and other functions” (p. 8). The alignment facilitates procurement execution, it converts into better performance (^{Rodríguez-Escobar & González-Benito, 2017}) and strengthens its recognition (Carr & Smeltzer, 1997). In fact, a growing comprehension of the strategic role of procurement in contributing to competitive advantage (Carr & Pearson, 2002) has drawn the attention of organizations to advance the alignment of corporate and procurement strategy.

Speed on decision-making is a critical issue for organizations and it is based on reliable and prevailing information. An organization must understand how to create, transform, and use the information to build a coherent view of a process, technology, and resource management (^{Choo, 1996}). However, information is becoming obsolete fast. Therefore, the faster an organization manages to use realtime information, the greater its conditions for making more assertive decisions (^{Eisenhardt, 1989}).

In this sense, big data concept is presented as a set of techniques and technologies that require new forms of integration to uncover large hidden values from large data sets that are diverse, complex, and of a massive scale (^{Hashem et. al, 2015}). It is not a matter of how to store or access data, but how to analyze it to make sense and exploit their value (^{Bello-Orgaz et al, 2016}). ^{Waller & Fawcett (2013}) intensify the idea of data science by coupling available and high-quality data and apply quantitative and qualitative methods to solve relevant problems in the field and forecast more assertive outcomes. ^{Provost and Fawcett (2013}) explain that data-driven decision is the way that one takes decisions based on data analysis rather than intuition or feelings.

The key challenge of procurement executives is how to deal with complex data, how to predict paths (^{Schoenherr & Speier-Pero, 2015}) and, consequently, how to move forward incorporating big data analytics strategies into procurement decision-making (^{Waller & Fawcett, 2013}). To support a superior appraisal of data dynamics and consequently leverage consistent decisions, organizations need to put efforts on the strengths of human and technology roles and for doing so, it demands a deep understanding of human judgment and technology capabilities (^{Schoemaker & Tetlock, 2017}).

Recent researches point to the relevance to advance and align data analytics strategies to the overall business strategy of the organization (e.g. ^{Agarwal & Dhar, 2014}; ^{Akter et al., 2016}; ^{Gunasekaran et al., 2017}). The alignment between corporate strategy, procurement strategy, and big data depends on the influence of top management (Gunasekaran et al., 2017) that synchronize interdepartmental competencies with the firm’s strategic objectives (Akter et al., 2016), resolve conflicts of interest and address the sense of urgency (^{Kayser at al., 2018}). In this sense, organizations need to develop an integrated model of analytics strategy that comprises the exploitation of internal data combined with external data sources, which includes environmental dynamics and its supply chain members’ information (^{Sanders, 2014}). However, according to ^{Castells (2011}), new technologies, when transforming the processes of information processing, act on all domains of human activity and enable the establishment of new connections between its elements and people. A virtuous circle of organizational transformation arises influencing the productivity and the efficiency of the institutions and consequently emerge new technological paradigms for the management.

It can also be observed in procurement dynamics. In line with Castells’ observations, ^{Sanders (2014}) affirms that, for the success of big data exploitation in the Supply chain (and Procurement as well), organizations need to reconfigure process, technology, and people behavior. It includes the need to comprehend the impact of individual values, behaviors, and social rules that underpin the current decision-making process. People, technology, and process are fundamental parts of procurement transformation. Firstly, information is recognized as raw material, while new technologies act on information. Secondly, while information is part of the entire human activity, processes are shaped by the new technological environment. Consequently, these dynamics demand a clear observation of the logics established by networks, upon the increasing complexity of interaction, speed, and variability above human cognitive comprehension of the phenomenon as well as human behavior (Castells, 2011).

Due to its complexity, big data cannot be used as a direct input on decision-making. It must be interpreted, and traditional tools should be insufficient to handle such a miscellaneous of data (Constantinou and Kallinikos, 2015). This perspective does not diminish the need for human value, on the contrary, it contributes to a better analysis of data in a broader way, which, possibly, a buyer would not have enough cognitive capacity to analyze such complexity of data interrelationships. Thus, organizations need capable people, in terms of programming, data analysis as well as the comprehension of organizational policies and business environment (Arkter et al., 2016; Carrillo, 2017; ^{Barbosa et al., 2018}). According to McAffee & Brynjolfsson (2012), teams tend to set decisions more clearly aligned to objectives, ask the right questions, and consequently are more able to bring the most assertive answers to the challenges posed. This knowledge allows people to clarify terms, issues and possible results gain support to undertake a speculative appraisal of data analytics (^{Braganza et al., 2017}), which supports organizations to plan and control strategies (Arkter et al., 2016) by creating a “trail of evidence” (Braganza et al., 2017), avoiding maverick buying. This approach implies a significant structure change (Barbosa et al., 2018) and consequently demands management commitment to support continuous improvements over time.

Hypothesis 1: Management alignment, knowledge, and coordination of technology strategies impact big data deployment on the procurement environment.

THE MICRO FOUNDATIONS OF EXECUTIVES’ BEHAVIOUR: INTENTION AND REACTION

From the emergent opportunity for adding value from the abundance of data, challenges for management arise. It includes individual cognitive, social, and motivational issues that are related to jeopardized decision-making (^{Van Knippenberg et al., 2015}). Scholars highlight the relevance of individual behavior during an intervention in operations management (^{Gino & Pisano, 2008}). ^{Boudreau et al. (2003}) mind to considering both technical and human aspects while investigating any changes in operating systems. It is discussed that the success of the implementation of tools and techniques in the supply chain field relies on human behavior. Individual behaviors, like lack of trust or risk aversion, tangle the initiatives to provide changes in operations management (^{Bendoly et al., 2006}).

Environmental characteristics, organization rules, internal demands, and procedures influence the purchaser’s behavior (^{Johnston & Lewin, 1996}), as well as, a background of individuals, information source, perception distortion, and satisfaction with past purchases (^{Sheth, 1973}). Purchasers are under significant stress to achieve targets (^{Giunipero & Pearcy, 2000}; ^{Tassabehji & Moorhouse, 2008}). When clear information lacks about the expectation associated with the order, the methods for fulfilling known procurement expectations, or the consequences of individual performance (Johnston & Lewin, 1996) procurers’ self-esteem and behavior are impacted. It is not rare to identify misunderstanding between purchasers’ role self-perception and the organization’s perception of its role (Tassabehji & Moorhouse, 2008). All these aspects drive procurers’ decision-making mental orientation for making a purchase.

Hypothesis 2: Procurers’ behaviors impact the deployment of big data analytics strategies on the procurement environment.

To better elucidate procurers’ behavior that should influence the deployment of data analytics, we exploit hypothesis 2 in a deeper detail based on behavioral assumptions intentions, actions, and reactions (^{Bendoly et al., 2005}).

RESEARCH METHODOLOGY

The research methodology adopted by this study is the survey (^{Forza, 2002}). The survey design comprises single-respondent research (^{Montabon et al., 2018}) who answer all items, including both the independent and dependent variables. It is a monadic study (focus on a single perspective). Once the core focus of this study is behavioral and drives this attention to the individual decision-making process, the potential for respondent bias is lower (^{Flynn et al., 2018}).

RESULTS

First, we sought to understand the characteristics of the survey sample. It’s constituted by 219 Procurement professionals, with the following scholarship: MBA (63,1%), bachelor’s degree (31,3%), Master (3,2%), High School (1,8%) and Ph.D. (0,5%). The executives are from several industries as Automotive (20,74%), Services (17,51%), Chemistry and Pharmaceutical (14,75%), Food (11,52%), Electronics (5,07%), Construction (2,76%), Energy (2,30%), Retailing (2,30%), others (23,05%). The sample is also compounded by 10% of companies with till 100 employees, 21% from 101 to 500 employees, 13% from 501 to 1000 employees, and 56% with more than 1000 employees.

Before testing the relationship between independent variables and dependent ones, we proceeded with preliminary data analysis (^{Forza, 2002}) by the dimensions refining, which is depicted in Table 1. We verified that some dimensions have the Cronbach Alpha lower than 0,70, however, the Composite of Reliability is higher than it. It is because the Cronbach significantly higher than 0,05 were eliminated from the analysis. After that, coefficients of reliabilities were calculated to verify the internal consistency of dimensions. It was used the Cronbach’ Alpha, Composite Reliability and Average Variance Extracted (AVE). The last one should be higher than 0,5 while the others higher than 0,7 to endorse internal consistency.

The dimensions were converted into variables by estimating their factorial score through the Promax extraction method. The variables were used to test the relationship between Big Data usage and Procurement Capabilities dimensions on Procurement Outcomes, using Linear Regression. It was considered significant the relations between independent and dependent variables in which the significance was lower than 0,05.

Alpha coefficient takes account of the sample size and number of items in a dimension. Thus, as the higher the number of items and sample size, the higher the reliability coefficient. In another way, the Composite Reliability was higher than 0,70 in all items, once it is calculated through the correlation among items instead of the number of questions in the dimensions (Table 1)(Table 3).

We subsequently estimated the factorial score of each dimension, that became independent variables, except the dimension “Outcomes” that is the dependent one. The regression analysis is depicted in Table 2.

According to Table 2, Knowledge and Strategy Alignment has a positive and significant influence on Procurement Outcomes (0,426). This influence suggests that the level of knowledge of Procurement staff upon programming, the comprehension of technological trends, strategic organizational plans, business environment, mission, vision, and firm strategy may improve the quality, compliance, and responsiveness of big data implementation in procurement. It also suggests that when a firm has objectives that are quantified, a detailed action plan, and prioritize investments on Big Data analysis, the Procurement Outcomes may be enhanced as well.

Dimension Planning and Control has also a positive and significant influence on Procurement Outcomes (0,381). In this dimension, it was assessed to what extent the organization examines the innovative opportunities to use the Big Data strategically and accomplish a plan to analyze the Big Data in a systematic way. The more activities are performed to plan and control the Big Dada, the higher the results from the Procurement department.

On another hand, Coordination has a negative and significant influence on Big data Procurement Outcomes. Although participating in meetings to discuss the procurement issues may generate better results, the frequency of the meeting may hamper the performance. It was perceived in the regression analysis due to the negative and significant relationship between Coordination and Procurement Outcomes (-0,159). Therefore, we suggest that the amount of information got from multifunctional areas in so many meetings may delay the decision making and, in consequence, harm the Procurement performance. In fact, these results offer significant pieces of evidence to confirm the first hypothesis, supporting the idea that management alignment, knowledge, and coordination of big data strategies impact its deployment on the procurement environment.

Considering the impact of procurement professionals’ behavior, the second stage of our study intends to verify if procurers behavior impact on the deployment of big data analytics strategies on the procurement environment. We split this hypothesis into 4 hypotheses exploiting procurers’ intentions and 5 hypotheses exploiting procurers’ common reactions and their impact on big data deployment within the procurement environment.

We start verifying procurer’s intention (price seeking, quality seeking, value-added seeking, and willingness to use new technologies). We noticed that the best price seeking seems to be relevant for advancing the procurement of big data strategies. The results present a positive influence on the decision-making process, more precisely, our findings suggest that it influences procurement in 12,2%. It reinforces the idea from previous research that most initiatives to implement complex data analysis in Procurement is motivated by elementary sourcing drivers such as cost reduction. Additionally, we verified that the data accessibility and the comprehension of it to make better decisions (willingness to use new technologies) influence positively Procurement Outcomes (0,143) and are also in the line of IT literature (^{Davis et al., 1989}; ^{Schoenherr & Speier-Pero, 2015}).

Surprisingly, Quality seeking and efforts dedicated to getting the best products in terms of quality have no influence on the use of big data in the procurement environment. Probably because quality has become a basic issue in procurement that the search for quality is a matter of qualifier criteria to buy rather than result or performance. There is no evidence that recognizing quality as related to price and choosing cheaper products rather than expensive ones’ influence on Procurement outcomes. It suggests that buying products duet to its price is not a matter of performance, but a choice.

In fact, we were outraged to not find enough pieces of evidence on the relationship between the search for value-adding and the usage of big data (0,054). It demonstrates the long journey that procurement executives have to exploit to understand the relevance of using complex data in Procurement. Apparently, interest is triggered by the mere possibility of reducing costs and it is not yet possible to understand its use to capture and generate value for the organization.

Besides, our results did not offer enough evidence to affirm that procurement professionals’ reactions due to daily routine pressures and current dilemmas impact the implementation of big data initiatives. In fact, the limitations that executives face because they have very rigid specifications or brand-specific targeting do not impact the decisions of whether to use complex data. It is the same when considering the autonomy of managers to make decisions. It is because managers don´t consider that getting an opinion about the procurement process from coworkers and bosses may improve the outcomes. The results also have shown that there is no evidence that buying from the favorite brand, keeping the same suppliers, and being loyal to them, enhance the Outcomes. The resistance to change suppliers even when there are better opportunities offered by others has no influence on Outcomes. Managers also declared that the over choice does not harm the outcomes. Therefore, we believe that multiple offers are not a fact that cause confusion and affect the outcomes due to the experience of managers in procurement areas.

To summarize, we realize that the dimensions that have more influence on Outcomes are related to Big Data, which means that procurement capabilities are elementary competences and not a differential, as expected. The combination of the five significant dimensions mentioned previously explain the Procurement Big Data Outcomes in 48,1%.

DISCUSSION AND CONCLUSIONS

In line with a continuous call for addressing cross-disciplinary research (Singhal & Singhal, 2002) our study combines consumer behavior, information technology, and procurement literature for better comprehend the role of organizational alignment on big data deployment on procurement activities and the impact of procurement managers behavior in the B2B environment.

Firstly, we attempt to evaluate the relationship between organizational, procurement, and big data strategies alignment. We found that several factors are relevant for procurement big data succeed: the technical and business capability of personnel involved, the preparation of the organization to support the implementation and use of big data, and the frequent inter-functional discussion of opportunities, plans, and responsibilities of data analysis. All these topics joined the knowledge, coordination, and planning of big data that should precede its deployment in procurement. Results suggest that without strong support and direction of C-level, big data is an agenda without the proper understanding of how to apply and the real worth for procurement. Future studies should address this gap, exploiting the levels of technology adoption, moving from pure digitalization to real big data expertise, and comparing to the level of alignment between corporate strategy and procurement strategy. Applied cases and roadmaps would be quite welcome to advance the comprehension of this phenomenon.

Secondly, we attempt to evaluate the impact of human behavior on the use and implementation of complex data analytics in the procurement environment. Our assumptions comprised that individuals must be capable and motivated to execute such actions and the environment shall offer adequate situations for development and motivation. Scholars emphasize the understanding of how an individual affects the system and outcomes (^{Boudreau et al., 2003}). Although behaviors are relevant in change management processes, the buyer’s current behaviors (intentions and reactions) do not significantly impact the deployment of big data strategies in this context, except “best price seeking” and “willingness to use new technologies”. This result seems to be coherent due to the traditional nature of the procurement role. The key driver of advancing on the more assertive decision-making process is to achieve superior results on cost optimization (Chick & Handfield, 2004; ^{Hong & Kwong, 2012}; ^{Lilien, 2016}), although the cost of implementing big data analytics is also a concern for practitioners (^{Van Knippenberg et al., 2015}; ^{Barbosa et al., 2018}). In addition, the study shows that the propensity to recognize the usefulness of complex data analysis as one of the significant dimensions that can contribute to the success of the implementation of big data into a procurement role. This positive behavior is a signal of technology acceptance and it is recognized as the first stage of assimilating a new technology and process change (^{Davis, 1989}; Karahanna et al., 1999). In fact, it reflects again a natural mindset of procurers, who intends to search for a cost-benefit outcome from every change.

On the other side, the absence of significance on other common buyers´ behaviors reframes the broad idea that behaviors are relevant in change management. Individuals immersed in constraint mindset as resistance to changes, lack of autonomy, confused by overchoicing or even the positive ones as the predisposition to pay more for quality, become secondary dimensions in the process. In fact, reactions are an individual response to a social change (Bendoly et al., 2016) and as soon as the circumstances change, reactions tend to be reshaped by the new intentions. For instance, as soon as a procurement manager will comprehend how to use big data, the overchoicing confusion should be minimized and he/she will feel more secure with the new methodologies in place. On the contrary, people transform themselves from new processes and technologies.

For this to occur effectively and to succeed in the implementation of big data on procurement roles and activities, corporate and procurement strategy must be in line. Once top management is convinced and sponsor the process change, current members’ individual behavior has a non-significant impact on the context. Moreover, a major implication of our findings for strategic sourcing and the challenges to implementing a big data-driven strategy is that knowledge and strategic alignment may serve as an important driver to support managers to comprehend how to include big data strategies among procurement decision making-process and consequently advance the value proposition of procurement role.

This study has also some limitations and researchers addressed a superior effort to mitigate them. The first issue refers to the methodology chosen and the informants’ approach. Our research was designed to inquire about a single respondent. However, some authors criticized it (e.g. ^{MacKenzie & Podsakoff, 2012}; ^{Roh et al., 2013}) arguing that it should trigger common method bias and potential respondent bias. We recognize the shortcomings of this decision and we put important efforts to mitigate potential risks for the inferences and results of this study, following literature guidelines. The first action in place was the commitment established between researchers and informants to assure the confidentiality and anonymity of research participants (^{Bell & Bryman, 2007}; ^{Wiles at al., 2008}). Although this strategy is the most appropriate for the informant, because it provides a certain comfort to answer the questions in a sincere way, this also made it very hard to guarantee that our database could be composed of multiple informants. We selected carefully the population and sample by focusing on procurement managers, who were knowledgeable practitioners (^{Montabon et., al, 2018}) involved (on daily basis) in procurement challenges and they can provide firsthand high-quality information of the phenomenon under investigation (^{Forza, 2002}; ^{Krause et., al, 2018}). Moreover, the researchers are committed to providing a summary report with the results of the research to informants. This strategy adopted during the searching and invitation of potential informants was a way of selecting people really interested in the contribution of the study to their own work and consequently avoiding distortions in the shared data (^{Phillips, 1981}).

Secondly, we increase our efforts to obtain responses from a substantial number of firms (^{Montabon et., al, 2018}). Since we surveyed a single respondent for all items, the potential for respondent biases reduces. This is justified by the fact that our study focused on behavior and individual decision-making process (^{Flynn et al., 2018}). Finally, even though that single respondent can be the Achilles’ heel of the present research, all constructs were previously validated, and the objectives, framework, and context are new for addressing the research question (Flynn et al., 2018).

This research provides preliminary shreds of evidence of the phenomenon and consequently, it offers the basis for more in-deep surveys, case studies, and experiments. For deeper comprehension, future studies should explore a longitudinal perspective comparing past constraints and how organization overcomes these challenges to adopt big data analytics. Additionally, we would like to see in near future studies that compare polar cases to comprehend the differences and convergences on the decision making the process for adopting big data in procurement.