It has been said that Big Data has applications at all levels of a business. This is definitely true of supply chain management – the optimization of a firm’s supply-side business activities, such as new product development, production, and product distribution, to maximize revenue, profits, and customer value. Big Data management has tremendous implications for supply chain management. Firms that can aggregate, filter, and analyze internal data, as well as external consumer and market data, can use the insights generated to optimize decision-making at all levels of the supply chain.

However, while many firms have noted the tremendous potential of Big Data for supply chain management yet not integrated it into their operations because they lack the financial, technological or human resources to do so. While these are clearly challenges, it is estimated that the digital universe will be over 40 trillion gigabytes by 2020 – a significant portion of that being data that can be leveraged to generate business insights. As time passes, those firms who have integrated Big Data into their supply chains, and both scale and refine that infrastructure will likely have a decisive competitive advantage over those that do not.

How to Optimize Supply Chain Management with Big Data

© | Mascha Tace

In this article, we will cover 1) the benefits of Big Data for supply chain management, including its role in 2) real-time delivery tracking, 3) optimized supplier chain management, 4) automatic product sourcing, 5) customized production and service, and 6) optimized pricing, as well as 7) building a Big Data supply chain, and 8) the future of Big Data and supply chain management.


By strengthening its supply chain, a firm can get the products and services a consumer wants to them quickly and efficiently. Firms that demonstrate such value to consumers can increase repeat purchase behavior, deepen consumer brand loyalty, and derive more value (purchases and referrals) from the customer over his or her lifetime.

To leverage this opportunity fully requires the firm to analyze internal and external data for decision-making efficiently. The management tools and techniques that have evolved for use with Big Data such as real-time business intelligence systems, data mining, and predictive analytics, can be leveraged to make fulfillment more efficient and profitable; optimize both supply costs and pricing to maximize profits; automate product sourcing; and deploy mass customization product strategies.


Big Data’s management systems include real-time analytics solutions that can be used to strengthen fulfillment. These systems include both Big Data hardware/software for warehousing and processing and inputs from bar-codes, radio frequency identification (RFID) tags, global positioning systems (GPS) devices, among others. Such systems can capture traffic sensor data, road network data, and vehicle data, in real-time to allow logistics managers the capacity to optimize delivery scheduling. They can address unforeseen events (such as accidents and inclement weather) effectively; track packages and vehicles in real-time no matter where they are; automate notices sent to customers in the event of a delay; and provide customers with real-time delivery status updates. Firms can also aggregate and filter relevant unstructured data from sources, such as social networking sites for insights on the delivery process, and respond to issues in real-time.

Further, vehicle sensor information can be used for predictive maintenance –maximizing the life of business equipment (in this case, vehicles and transportation-related equipment such as forklifts) by scheduling preventive maintenance based on current and historical data.

Transportation data, when integrated into a commercial or in-house implementation of a distributed file system, such as Hadoop, a network-based one like Gluster, or other similar system, can be leveraged by other strategic business units. For example, a firm can configure its transportation business intelligence system to route notification of delivery delays to customer service centers automatically; customer service representatives can then anticipate, and respond to, customer complaints appropriately.


To maximize profits, firms want to sell the most products at the lowest costs. Cost determinations become increasingly complex the more raw materials used to produce a product, the greater the variability in the price of those inputs, the more products the firm offers, and the larger the geographical distribution area. The supplier relationship management process – which once, for many firms, had more to do with drinks, golf games, and other shared social experiences – these days, must incorporate more quantitative measures to determine whether the firm is receiving the most bang for its buck.

Big Data allows firms to develop complex mathematical models that forecast margins if different mixes of suppliers are chosen. These models can take into account a wide range of variables, such as the additional costs due to variations in the speed with which different suppliers can deliver their goods; one-time switching costs, such as long-term contract cancellations; and even estimates of supplier reliability, which firms can use to generate performance predictions of various supplier mixes. Managers can then select those with the highest return on the lowest investment to maximize profits.


Similar to supplier selection, Big Data has many benefits for pricing. Firms can use consumer data, from both internal and external sources, to develop pricing models that maximize profit margins, and use predictive analytics tools to forecast demand for a particular product at different price points. Firms can then test these price points with soft launches, and incorporate consumer behavior and feedback – both quantitative and qualitative – into their pricing strategies. Further, firms can develop models to determine which combinations of related products consumers are likely to buy together, and use this information to develop and refine upselling strategies.

Another application of Big Data management and analysis to pricing involves sales forecasting. Firms can use predictive analytics to make real-time predictions about the firm’s sales performance overall, in a region, or even a specific location; they can adjust pricing to ensure that they meet those projections when necessary. Dynamic pricing can also be used to maximize revenue during times of increased market demand and/or supply shortages. Common in ground and air transportation during the holidays, dynamic pricing allows operators to increase prices for empty bus, plane, and train tickets when empty seats are scarce. However, industries ranging from hotels to sports entertainment to retail employ dynamic pricing to increase revenue.


Big Data collected to optimize supply chain management often holds key insights about consumer needs and wants. Firms can leverage these insights to develop new product and/or brand extensions, where sufficient consumer demand warrants. In many cases, economies of scale reduce the costs of product extensions to the point where the additional costs are negligible. For example, a firm might introduce a jacket in three different colors, but through an analysis of aggregated social media mentions, customer service feedback, and online reviews, release the product in a fourth color. This is known as cosmetic customization.

Many firms also leverage economies of scale to employ a mass customization strategy – one where customers provide firms with product features for common products, and the firm builds the product to the customer’s specifications. Auto manufacturers often employ this strategy, manufacturing large volumes of common components, and then allowing users to “build” their car by inputting desired features on the corporate website. However, many firms, from eyewear designers to toy companies, use this strategy, known as collaborative customization.

Other firms, such as software firms, employ adaptive customization, which provides users with products that consumers can then customize themselves, according to their changing needs and desires. Still others employ transparent customization, wherein customers do not know that firms have customized products specifically for them. Often, this is employed not only with product manufacturing but also with fulfillment: firms analyze consumers’ usage patterns of commodities, and produce and offer, and distribute replacements when needed.

In addition to adding value for the consumer, mass customization enhances a personalized purchase experience considerably, deepening both brand engagement and loyalty. Firms often use Big Data, including supply chain data to personalize their customer service experience. Firms with effective customer service departments integrate all available data about a consumer, including relevant supply chain data (such as a history of on-time and delayed deliveries, for example) into files available to customer service representatives. Having that data at their fingertips helps customer service reps address customer inquiries received.

Firms can even use this data to anticipate such inquiries and respond proactively. For example, a firm might face greater demand for a particular product than they have inventory to meet. In such a case where the product has a lengthy manufacturing and/or distribution time, the firm can reach out to those who have placed orders with an explanation and apology for the delay; they can also update their website to notify new customers of the delay.


In late 2013, Amazon filed a patent in the U.S. for the process of predictive shipping – a distribution method wherein a firm uses predictive analytics to forecast future sales based on historical data; they then source and ship products to local and/or regional distribution centers in advance of those orders. It remains to be seen how successful this method may be, yet given Amazon’s pioneering success in the online retail space, driven in no small part by its embrace of Big Data management tools, techniques and technologies, it would be tough to bet against them.

Twelve years earlier, the firm filed a patent for automated product sourcing– a process and its related technologies that played no small part in Amazon’s success; it has since been replicated by many other online retailers to varying degrees of success. Automated process sourcing refers to a firm’s ability to, upon receipt of a customer order, analyze inventory at multiple fulfillment centers, estimate delivery times, and return multiple delivery options (at different price points) to the customer in real-time. This enhances value for the customer, and allows Amazon to optimize distribution, as well as inventory management. Many other firms, from Best Buy to eBay, have either developed their own automated product sourcing systems or purchased software and process management solutions from vendors.


The benefits of paring Big Data with supply chain management make it an obvious choice; the ever-accelerating volume, velocity, and variety of data make it a necessary one. However, integrating Big Data into a firm’s supply chain is more involved than releasing a management directive or signing a purchase order.

It is often advisable to start with individual links on the supply chain – such as departments, build Big Data into their operations, and replicate their successes across the organizations. The buy-in from this approach will help managers mitigate internal resistance to an innovation many find abstract or overwhelming. Executives and managers must review (and where needed update) the strategic business goals that drive the specific operational unit.

For example, a corporate fleet might count as KPIs on-time deliveries, cost per delivery measured in fuel, wear and tear, and other measures, delivery times, positive customer feedback, lack of negative customer feedback, and other similar indicators. Internal data scientist leads should work with must work with executives and managers (in this case the management team of the corporate fleet) to create operational goals and insights that drive these goals. For example, such insights might include the optimal time by which deliveries must be made to elicit positive customer feedback, optimal delivery routes that minimize cost per delivery and delivery times in real-time, and others that can allow the corporate fleet to add value to the organization as a whole. Data scientists then must work with I.T. (and vendors where necessary) to develop a Big Data infrastructure that allows them to meet these goals.

Fundamentally, such architecture would include hardware/software and internal procedures and protocols for collecting, processing, and storing existing and new data, in real-time where possible and necessary. This architecture would also allow data scientists to clean, search, and filter data pre-analysis, analyze it as necessary, generate useful reports, and share actionable insights across the organization, and in some cases, to consumers. Further, this architecture must be scalable – as the volume of data will only grow, and secure, as a failure to maintain the privacy of consumer data can be a tremendously expensive mistake. Such architecture should communicate with existing (or new) customer relationship management systems and provide real-time intelligence to provide the most value for internal and external stakeholders.


Several innovations and trends will not only accelerate the volume of data as a whole, but also the volume of data relevant to supply chain management. Mobile will continue to provide a major source of supply-chain relevant data, driven by the GPS technology in mobile devices, as well as the proliferation of social networks specializing in social discovery, which allows users to discover people and events of interest based on location. Deep analysis of consumer location information can afford firms even greater efficiency at getting products to consumers, whether through optimizing the locations of regional fulfillment centers or even distribution of products at those events and venues well frequented by its consumers.

The Internet of Things – the attachment of sensors and other digital technologies to traditionally non-digital products to capture data, are currently, and will continue to be a major source of data of use to data scientists working on supply chain optimization. For example, a smart device can be built to send messages to the manufacturer when they are broken, which can generate production on a replacement part or full device, before its owner calls customer service. If the device is outmoded, its signal to the manufacturing firm can provide the customer service representative (and/or sales staff) with the information to prepare for an upsell.

Cloud computing itself has driven Big Data’s growth significantly, as its inherent digitization of a firm’s operational data demands new methods to leverage it. As more firms take advantage of the benefits of cloud computing (such as reduced capital costs, economies of scale, and increased flexibility), adoption of Big Data’s management tools and techniques will grow. Moreover, as it grows, firms will demand increasingly sophisticated business intelligence systems, methods of predictive analysis, and tools for data mining, which the market will provide.

Comments are closed.