More data crosses the internet every second than was stored in the entire internet just 20 years ago. With social networks, sensors, cameras and smart phones, we are now walking data generators.
Big data refers to the massive amounts of data captured by point-of-sale systems, sensors, web server logs, internet clicks, social media activity reports and smart phone records, every second. With this almost overwhelming amount of data comes a need for a new way to process all that information into something useful.
“Big data analytics gives us the tools to examine big data using advanced technologies to provide data management, open-source programming, statistical analysis, visualization tools, in-memory computing and more,” says Dr. Morgan Swink, the Eunice and James West Chair of Supply Chain Management and director of the Supply and Value Chain Center at TCU’s Neeley School of Business.
More than simply gathering information, analytics turns big data into a powerful tool for forecasting/demand management, logistics improvements, inventory optimization, network design/optimization, CRM/customer/patient analysis, purchasing spend analytics, production run optimization, warehouse operations improvement, process/equipment monitoring and sourcing analysis.
Both large and small companies utilize big data, with B2C industries among the major users.
Yet big data is largely going untapped. Why? Swink sees six major hurdles:
Lack of Vision/Business Case for Benefits
Companies thriving on big data start with a targeted approach. Swink cites Sears as an example.
“Sears uses big data to influence its pricing strategy,” Swink says. “At one time Sears followed a nationwide pricing strategy, which was then reduced to a regional pricing strategy. Today, with approximately 4,000 stores and more than 100 million customers delivering a stream of big data, Sears’ objective is to deliver personalized pricing and offers.”
Similarly, Swink points to Wal-Mart, which uses big data for its mobile marketing strategy.
“Customers who use the Wal-Mart app spend 40 percent more per month at Wal-Mart than customers who do not,” he said.
According to Gibu Thomas, global head of Wal-Mart’s Mobile Division, Wal-Mart is leveraging big data to develop predictive capabilities to automatically generate a shopping list for customers based on what they and others purchase each week.
Lack of Visibility/Access to Data
“In the supply chain, we already have several measured elements of visibility,” Swink says. “From customers we can derive, share and track sales, demand forecasts, inventory levels and promotional plans. From suppliers, we can monitor inventory levels, lead times, delivery and advanced shipment notices, as well as check on the status and location of finished goods. Other systems provide market-level demand and market-level supply information.”
But how accurate is this data? How timely is it? And does it come to us in an easily used format?
Swink believes that, to make real impacts, big data analytics must deliver precise information, available almost immediately, in usable formats that generate visible results.
Lack of IT Infrastructure
According to Swink, only about 25 percent of big data users say that they have the mobile technology capabilities needed to deploy the information. And that’s among the heavy users. Among the low users of big data, the percentage drops below 20 percent.
Among the high users, the highest use percentage f use comes from tracking assets, checking process status and conducting operational transactions remotely. The top hindrance to low users is the lack of ability to conduct operational transactions remotely.
“Acting on big data requires execution technologies,” Swink says. “The biggest users of big data are also well-versed in execution systems (TMS, WMS), ERP, advanced forecasting/S&OP, advanced planning (APS, APO), and to a lesser extent, SRM and CRM. ERP is the predominant technology for low users of big data. SRM is the weak point for low users.”
Lack of Analytics Capabilities
Swink says those who use big data regularly deploy dashboard applications, data visualization techniques and advanced analytical techniques that combine and integrate information.
“Low users have the most competency in data visualization techniques, but they tend to lack systems that automatically make operational changes,” he says.
Lack of Organizational Structure
The effectiveness of big data and data science is moderated by domain knowledge. A company must not only collect information, it must decide how to best use it,” Swink says. For collected information to be useful for decision making, it must be available to managers who have relevant business knowledge.
“So, a key to the effective use of big data is the company’s level of internal (cross-functional) integration,” Swink says. “Internal integration allows the information to flow quickly to the right decision-maker and aligns the information needs of the company with the business processes.”
Internal integration is vital to sharing and using big data.
“Companies should ensure that functional teams are aware of each other’s responsibilities, goals, metrics and data sources,” Swink says. “They should develop integrated planning and a common prioritization of customers’ needs across functional teams. And they should facilitate the regular exchange of operational and tactical information between functional teams.”
Big Data is Yours for the Taking
Even with the above challenges, there is still plenty of time to stake out a leadership position for a competitive advantage.
“Success requires a clear vision and business case coupled with complementary assets such as a supporting technological infrastructure (connectivity/visibility, mobile deployment, and systems for analysis and execution), analytics capabilities and supporting integrated organizational structure,” Swink says.