Introduction

Big data has become one of the most important aspects of supply chain management. The concept of big data refers to the massive data sets that are generated when millions of individual activities are tracked. These data sets are processed to yield insights that help inform managerial decision-making. Supply chains in particular have leveraged big data because companies have been able to develop technology to not only capture hundreds of millions of data points, but to process them in meaningful ways to eliminate waste and promote efficiency in the supply chain systems. This paper will examine the concept of big data, how it has arisen and come to dominate supply chain management, and look at the different ways big data is transforming the supply chain function. Lastly, the paper will take a closer look at the future for big data with respect to supply chain management. As it becomes easier to gather data, and as there are diminishing returns to statistical robustness as the number of data points increases, are the competitive advantages of big data going to diminish?

The Evolution of Supply Chain Management

The field of logistics management was focused on controlling the flow of materials, in-process inventory and finished goods through a company's system from the time that it enters the system until the time that it leaves the system (Cooper, Lambert & Pagh, 1997). As the field became more strategic in nature, it came to encompass other issues, such as sourcing materials and building in redundancy (Cooper & Ellram,1993). More than simply moving things from point A to point B, the field became holistic in nature, where the quality and price of goods were factored into purchasing decisions as well as the logistics of getting those goods to the right place at the right time. Driving this change was the move towards a globalized marketplace. Globalization increased the complexity of the supply chain, adding longer transportation routes, border wait times, currency exchange, duties and tariffs, and a host of other variables that now had to be taken into consideration logistics has remained important but it always viewed in context with the rest of the supply chain.

Big Data

The concept of big data really began to arise in the 1990s but has become increasingly important since that point. Big Data refers to the use of very large data sets to enhance managerial decision-making. The concept of big data arose as technology has developed to allow businesses to capture enormous data sets, and process them relatively easily (Boyd & Crawford, 2012). Companies have long collected data at a rudimentary level. Loyalty programs and credit cards represented an evolution in the ability of companies to collect data and distill that data into consumer spending habits. This information is then made actionable by letting companies understand more about buying patterns. Big data is similar, but with a lot more data. One of the major advantages of big data is that it allows for complex problems to be solved. A modern supply chain can be exceptionally complex, and one of the important things about this complexity is that no one person can effectively make all the decisions decision-making tools are needed that can ensure not only consistent decision-making across the company but coordinated decision-making as well (Hult, Ketchen & Slater, 2004). It is these coordinating mechanisms where the true power of big data lies being able to identify things and make decisions that an entire team of humans working without big data would probably never be able to identify (Fugate, Sahin & Mentzer,2005). Once big data gets to that point, a company can generate true competitive advantage. And when a company is large enough that is has a data advantage, it will be able to sustain that advantage, which is why there has been such a rush in recent years with respect to big data.

As the concept was being fleshed out in academia, businesses were just starting to learn what they could do with all of the information that they were collecting and one of the applications was to move away from marketing and use data to make decisions about the supply chain (McAfee & Bryjolfsson, 2012).

One of the first steps that companies needed to make was to hire data scientists the sort of people who could process these data sets and derive useful information about them. Data scientists suddenly became popular, for their ability to take vast... At the heart of the drive to adopt big data is competitive advantage. Companies have invested in their data programs because they can derive significant advantage from big data under two conditions. The first is that larger companies have access to more data than smaller companies. The incremental cost of data acquisition is lower, and the company's ability to use that data in decision-making is theoretically better. The second is that even among larger companies, there are first-mover advantages to be had. This is evident in the supply chain, especially among companies that are competing on price. Using the classic example of Wal-Mart, one of the leaders of data-driven supply chains, the company competes on offering the lowest prices, as do most of its competitors. Thus, if it can lower the cost of getting goods to its stores, it can pass those savings along to customers. There is opportunity for competitive advantage under that scenario, if cost leadership is the chosen strategy. Even when cost leadership is not the strategy, making the groundbreaking decision early puts a company in a better competitive position than its competitors (LaValle, et al, 2010).

Big Data in the Supply Chain

As the largest non-oil company in the world, Wal-Mart is looked to as a leader, so the fact that they were first movers on the use of big data in supply chain management has ensured that the rest of retail and other industries as well have followed. Some of the technologies that Wal-Mart has adopted allow the company to track its inventory from when it leaves the supplier if not before all the way through the logistics channel. Once Wal-Mart takes possession of the good, that good is scanned regularly through the process. The company's trucks are tracked via satellite. Stores use automatic re-ordering triggers to ensure that goods can be received as soon as they are needed. The goals of all this are to lower inventory holding costs by reducing the amount of inventory that stores have. Goods are turned over more quickly, because Wal-Mart receives them only days before it expects to sell them. Big data plays a significant role in ensuring that this process can be achieved. There are a couple of key areas highlighted for big data in supply chain management.

Demirkan & Delen (2013) note that data, and how a company uses its data, is one of the ways it can truly differentiate from its competitors. It can be difficult to truly and consistently attract superior talent, and it can take time to move the needle on brand image, but data has become a popular means of finding competitive advantage largely because it is new, and firms in many industries are basically in a data arms race to find innovative ways to use their data to extract competitive advantage.

The first is predictive analytics. Data science often focuses on using past events to predict future ones, and that is one of the main uses for big data in supply chain management. For example, if Wal-Mart in Smalltown, OH is running out of shovels at the end of February, and it takes twenty days to order new ones from China, including manufacturing and shipping times, three things can happen. The company can order a lot of shovels and ensure that they have supply. If spring comes, those shovels will sit in a warehouse until next November. They could also run out of shovels, but a late-season snow could leave demand on the table if the store lacks inventory. Modelling both weather patterns and loca buying patterns can help the company to settle on demand. Even when weather is not a factor, the company can examine past purchasing patterns to set order quantities. The earlier it can set these quantities, the better response it can get from suppliers. Wal-Mart knows already what the normal amount of hot dogs it sells on the 4th of July, for example, so it can feed that information to its suppliers to ensure that they have those dogs at the Wal-Mart warehouse, exactly in the quantity Wal-Mart needs.

Predictive analytics is used in supply chain management to take the variability out of the system as much as possible. Inventory usage is reduced, as is the potential for waste, especially with perishable goods. The chances of disappointed customers is also reduced. It is almost impossible and certainly it is impossible for a company like Wal-Mart to have exactly everything delivered exactly when the customer…