Procurement analytics, which supports decision making in procurement management, typically handles problems and decisions that are related to cost reduction, supplier management, cost modeling, procurement-led innovation, market strategies, supply chain risk, and stakeholder value improvement (Handfield, Jeong, and Choi 2019). As the domain of procurement management expands and the enabling technology for procurement analytics evolves, the scope of procurement analytics will also expand. We can think of this expansion in terms of organizational scope and temporal scope.
The organizational scope of procurement analytics mirrors the domain of procurement management and grows as the role and perception of procurement management expand. In the early era of procurement management, the main role of procurement management was to contribute to achieving high efficiency in internal resource management. Despite a few pioneering perspectives in 1950s and 1960s (Alfalla-Luque and Medina-López 2009), supply chain management (OSCM) centered on operations within internal organizations during its early history. Thus, the focus of procurement management in this era was chiefly on how to efficiently trace and utilize tangible and intangible goods purchased from suppliers. Thus, procurement analytics in this scope typically worked as a part of resource planning systems, digitizing and managing the information of resource inputs. While this role of procurement analytics improves the efficiency of internal operations (e.g., better order timing), the digitized information typically entails information quality issues. Information within organizations is often managed with non-comparable formats and incoherent schema, lowering the veracity of information and resource efficiency . Even today, emerging research finds that managers observe the non-negligible level of inaccuracy in data for physical goods challenges and that such inaccuracy requires time-consuming practices to be corrected.
The organizational scope of procurement analytics shifted from efficiency of internal organizations to information interfaces with suppliers as the importance of outsourcing increased. In addition, the wide introduction of IT-enabled digital communication between organizations lowered transaction costs by integrating supply chain in the information domain. This inter-organizational information flow entails new challenges in procurement analytics: organizational boundaries generate information friction and lowers the visibility of information (Barratt and Oke 2007). Since the level of data shifts from internal information to inter-firm information, dyadic level data management, which has higher technological complexity and coordination cost than management of internal data, is required. Moreover, sourcing organizations increasingly adopted centralized sourcing strategy, raising the importance of data analytics in sourcing. Since both outsourcing and centralized sourcing are primarily conducted to reduce costs , procurement analytics is expected to provide groundwork that assists managers with decision making within alternative sets (e.g., visualization and decision support). Furthermore, the information quality issue is still present in this scope: information distortion determines the value of information sharing across organizations in buyer-supplier relationships Barratt and Oke 2007).
The organizational scope of procurement analytics was also extended to suppliers beyond first-tier suppliers (defined as extended suppliers, hereafter), as the conceptualization of the boundaries of suppliers in procurement management began to include these “hidden” suppliers in the global supply chain. This drove sourcing organizations to pay more attention to complexity, risk, and uncertainty in their supply bases than the previous era. Yet, increasing the responsiveness of supply chain to address these issues can be easily constrained by the high degree of complexity in identifying extended suppliers and low information visibility in controlling and monitoring them. Because of the deficiency of direct information channels to extended suppliers and their circumstance, sourcing organizations can be incentivized to employ Internet-enabled external data. However, due to limitation in information access, external data will begin to source various information including information based on unstructured data (e.g., news text and government statement) to obtain comprehensive understanding of their supply chains at high cost of data processing and insight generation.
The temporal scope of procurement analytics pertains to the timeframe on which insights generated from procurement analytics are primarily based. The most basic temporal scope is a set of previous events. In contemporary organizations, managers derive principles from evidence and translates them into practices to solve problems. This “evidence-based management” (Rousseau 2006), by definition, requires a clear proof for a certain alternative. In procurement management, historical transaction data can help managers obtain clues to optimal solutions, especially when problems to solve result from repetitive and standardized processes. Since such data is typically first-hand data generated during operations, the accessibility of the data is high. However, this “evidence-based decision making” is pragmatic only under the premise that sourcing organizations are able to process necessary data for evidence generation. In the age of “big data”, the volume of data can be enormous, making it difficult for sourcing organizations to handle it without proper tools even if the quality of data is guaranteed.
The advanced scope of procurement analytics will move beyond current events. This scope overcomes the weakness of procurement analytics based on historical data; after all, there is always a gap between past and present patterns of observations. The increase in computing power (e.g., cloud computing) has compressed the required time for data processing and gave rise to the birth of real-time data analytics. However, as the volume of data grows, simultaneity of observation and analysis become a requirement , but exposes the risk of trade-offs between the reliability and depth of information with data processing speed.
The most sophisticated temporal scope of procurement analytics deals with upcoming events. This approach provides a solution to the key limitation of the aforementioned temporal scopes: when taking managerial actions based on past and current observations, managers cannot eliminate a lead time between the observations and the actions. Predicting upcoming events not only reduces risk and uncertainty that managers perceive during business operations but also narrows the time-window between the observations and the actions. Thus, this approach helps increase the responsiveness of a supply chain and secure competitive advantage by preemption of related resources. Yet, the inherent nature of prediction hinders the active implementation of this approach. Of all, the biggest challenge is that observations of upcoming events, by definition, are not yet to be revealed. Therefore, procurement analytics will need to develop “intelligence” that runs upon not only previous observations but also insights gleaned from understanding of ongoing trends and circumstances related to subject matters. This process necessitates use of diverse qualitative information sources that provide contextual clues and thus consist of a large portion of unstructured data. The role of intelligence is not limited to the prediction. After all, the primary purpose of addressing upcoming events is to shorten the lead time to the action. The intelligence of procurement analytics needs to consider alternative sets and determine the optimal choice within the sets. We may see more of this in the future.