Supply chain woes continue to plague organizations around the world and in virtually all sectors. For some, leveraging data and analytics tools is proving to be an effective way to address the challenges.
Disruptions to global supply chains due to the COVID-19 pandemic have been significant. As consulting firm Deloitte notes, the free movement and operation of people, raw materials, finished goods, and factory operations have been stymied. “Direct supply chains have experienced challenges, and so have extended supply chain partners such as third-party and fourth-party vendors — the suppliers of suppliers,” the firm says.
Enterprises face multiple risks throughout their supply chains, Deloitte says, including shortened product life cycles and rapidly changing consumer preferences; increasing volatility and availability of resources; heightened regulatory enforcement and noncompliance penalties; and shifting economic landscapes with significant supplier consolidation.
Technology can’t resolve every supply chain issue. Goods need to be produced and moved from point to point. But the latest analytics tools, powered by machine learning algorithms, can help companies predict demand more effectively, enabling them to adjust production and shipping operations.
Here’s how three organizations are succeeding at using data analytics to improve supply chain operations.
Enhancing operations and relationships with suppliers
Optimas Solutions, a manufacturer and distributor of fasteners, is using data analytics in three critical areas to improve operations and relationships with its suppliers and customers, says Mark Korba, vice president of supply chain and business intelligence at the company.
First, Optimas is using data analytics internally for a number of functions, including material acquisition for manufacturing; forecasting of production and customer demand; improving efficiency and accuracy with ordering from suppliers; and managing its inventory.
This has all helped Optimas manage and reduce overall costs by enabling it to make smarter decisions, “making our supply chains more efficient and improving overall cash management,” Korba says.
Second, Optimas is using data analytics to help better collaborate with its business customers to reduce costs and better manage their inventories. Analytics is also helping the company better predict demand and consumption. “By being able to perform these types of analyses it consistently helps to reduce costs,” Korba says.
Finally, Optimas uses analytics to better collaborate with suppliers. “By understanding and managing demand, especially individual customer demand, we provide more accurate forecasting data to suppliers and better manage our orders so they can work more efficiently for us,” Korba says.
The company is using a platform called Service Optimizer 99+ from ToolsGroup for demand planning, inventory optimization, and replenishment planning. The platform integrates well with Optimas’ NetSuite enterprise resource planning (ERP) suite to leverage supply chain data, Korba says.
“Often people think of the supply chain as one thing and it is not,” Korba says. “We think of the supply chain as the sum of several parts of the whole business operation — from understanding customer demand to materials management and manufacturing or sourcing and purchasing, to logistics and transportation, to inventory management and automated replenishment orders at Optimas and at our customers’ locations.”
A key to success is the ability for all the supply chain tools the company uses to work together seamlessly, to help keep customers appropriately stocked and better manage costs, demand, inventory, production, and suppliers. The information provided through analytics needs to address financial issues such as cashflow and pricing on the supply and demand sides.
“Overall, the supply chain problems all the tools address — whether working together or individually — improve efficiency, customer inventory management accuracy, supplier relations, cost savings, and the ability to forecast more accurately and quickly,” Korba says. “Data analytics has helped us gain visibility throughout the organization, even in places we never imagined such as better accuracy with our cash conversion cycle, ranking supplier responsiveness, analyzing time to task, or evaluating capacity of both our customers and Optimas.”
Supply chain data doesn’t necessarily have to reside at any one location, Korba says. “It is important to understand what systems or outside services can help you gather and analyze the data you need, so it can become useful decision-making information,” he says.
For example, using price indexes for materials such as steel and packaging or labor and transportation costs are helpful to determine when price increases are required. “Better access to information, inside or outside the organization, enables better decision making for Optimas and our customers and suppliers,” Korba says.
As Optimas has become more diligent about data analytics, “our customers and suppliers benefit from marked improvements in a number of areas, including better visibility into the entire supply chain,” Korba says.
Predicting product demand and inventory needs more effectively
CarParts.com, which sells automotive parts online, is using advanced data analytics tools and machine learning algorithms to better predict product demand and inventory needs.
“New data analytics capabilities allow us to be more accurate when forecasting demand for each city [and] region of the country,” says Stanislav Tatarzuk, vice president of inventory planning and forecast.
The company uses models derived from machine learning to figure out where to place inventory in its distribution center network, which allows it to get products closer to the customers who want them and to be able to deliver parts faster while also saving on shipping costs.
CarParts deployed a platform from Databricks that enables it to centralize all the data related to product orders and inventory and to apply analytics to the data. It is also leveraging open source libraries such as XGBoost and Prophet; a variety of analytics tools; and Apache Airflow, an open-source workflow management platform, for data engineering and report automation.
One of the benefits of using analytics to better manage the supply chain is shipping optimization. “The questions of where to ship from and how to consolidate into one box are complex,” Tatarzuk says. “Using live data from our shipping partners as well as our warehouses allows us to be as efficient as possible while using advanced analytics to optimize multi-item order shipping.”
That enables the company to get items to customers more quickly, while reducing shipping expenses. Using this process allowed CarParts to realize significant savings, Tatarzuk says, although he didn’t provide specifics.
Another benefit is warehouse optimization. “Being able to cut on labor expenses and shipment time is an absolute must in today’s wage-raising environment,” Tatarzuk says. “We are creating models to correctly place inventory throughout our distribution centers, to cut down on picking and put away time while creating high-density clusters that shorten pick speeds.”
Data analytics has helped the company deal with the disruptions in supply chains caused by the pandemic, by enabling it to see changes in demand and increasing lead times in the early stages of the supply chain crisis, and to react faster than its competitors.
Supporting an increasingly complex supply chain
Global technology provider Lenovo has been addressing the challenges of its global supply chain due to the pandemic by leveraging advanced forecasting technology and data analytics, says Arthur Hu, senior vice president and CIO.
Lenovo’s supply chain once focused primarily on logistics, information flow, and business flow, Hu says. But the company’s transformation into a full-service technology provider “has meant that our supply chain, once focused primarily on devices, has become increasingly complex, with more diversified customer demands, more complex products, and the need for more efficient and agile operations and service,” he says.
In the past year, the supply chain team has worked with 2,000 suppliers to deliver more than 130 million Lenovo devices.
Given the shift, the company’s supply chain team decided to revamp its operations, taking an “intelligent transformation” approach. “A cross-functional team worked to transform Lenovo’s supply chain operations into a data-driven, intelligent ecosystem,” Hu says. “The new system provides real-time data, intelligent analysis and decision-making support that allow our businesses to deliver on their promises to customers more effectively and efficiently than ever before.”
The company built a Cost Forecasting Engine (CFE) system to provide faster and more accurate forecasting for procurement, manufacturing, and sales costs throughout its supply chain operations.
Using the system in combination with linear regression and XGBoost (eXtreme Gradient Boosting), an open-source software library that acts as a machine learning algorithm, Lenovo’s managers can establish the maximum and minimum threshold to avoid extremes that affect cost accuracy.
The technology can make cost comparisons to reduce the impact of month-to-month cost fluctuations for hardware components, and provide a basis for managers to make business strategy decisions, Hu says.
The CFE now supports procurement and production cost-forecasting for more than 70% of Lenovo’s entire global supply chain, Hu says, and cost-of-sale forecasting for more than 90% of the supply chain. Compared with manual cost maintenance, cycle cost-forecasting efficiency has improved by about 12%. The cost-accuracy rate remains about 95%, he says.