The term “big data” is the buzz word du jour. But to drive the opportunity, companies must bypass the hype.
New forms of data do not fit traditional IT architectures. Traditional supply chains were architected to use structured data using software based on relational databases. The big data era will make many of the investments obsolete from the last decade. So, is it a problem or an opportunity?
At Supply Chain Insights, we think, and 76 percent of respondents on our recent survey agree, that big data is an opportunity. (See Figure 1, below.) Big data offers new means for the corporation to listen, test and respond faster, but only if enterprises change their process paradigms.
In this recent study, companies see the greatest opportunity to use big data for “demand” (to better know the customer and improve the response). However, actual current investments are in the areas of “supply” not “demand.” Respondents view supply-centric projects, like product traceability (involving product serialization and real-time location information), supply chain visibility and temperature controlled handling of goods using sensors as important and immediate investment areas.
In a recent conversation, a senior supply chain leader commented, “We have done what we can to make the supply chain efficient. It is now time to redefine demand.” Our response is, “Not so fast.” New forms of demand data, and the evolution of demand analytics, will result in the redefinition of supply-centered applications. That makes most of the architectures of the last decade legacy systems. As new forms of data emerge and best-of-breed solutions for analytics appear, corporations need to stabilize enterprise resource planning (ERP), or system-of-record investments, and begin to redefine the processes of supply—sell, make, source and deliver—based on new forms of data. Instead of looking outward from the enterprise, it’s time to look outside-in.
For the purposes of this article, we define big data as data with a volume greater than a petabyte coupled with a growing variety of data. A petabyte of data sounds like a lot of data; but, how much is it really? A petabyte of data is 1024 terabytes. A terabyte of data is 1024 gigabytes. In more graphical terms, a petabyte of data represents 20 million four-drawer filing cabinets filled with text or the storage of 13.3 years of HD-TV video. Twenty petabytes represents the amount of data processed by Google on a daily basis. It also represents the number of hard disk drive spaces manufactured in 1995. For 15 percent of manufacturers it represents the current size of their ERP databases.
New concepts are emerging to use these larger datasets. The greatest opportunity comes with the use of a variety of data.
It’s Still Early
Most companies are at the starting line in their understanding of the opportunity of implementing big data with their supply chains. Here are some additional insights from the survey:
- Enterprises need to get line-of-business leaders involved. The IT group is more likely to see big data as a problem, not an opportunity. Almost half—49 percent of those with a big data initiative—report that it is headed by an IT leader. These processes are the most successful when led by a line-of-business leader.
- The big data journey has just begun. Only 28 percent of companies have a big data initiative today. Only 15 percent of companies have a database that holds more than a petabyte. The largest databases are in two areas: demand insights for consumer products and product information for discrete manufacturers.
- Data-driven companies are ahead of the pack. Companies with strong master data management core competencies are further along on their understanding of big data and are more likely to have a cross-functional team studying big data.
Going After the Prize
While success in a big data supply chain strategy can ignite new business models and drive channel opportunities, the effort must be about more than the data. Seizing this opportunity requires leadership and the initiatives a corporation pursues need to be aligned to business objectives, with a focus on small and iterative projects. Here are five recommendations:
1. Build a cross-functional team that focuses end-to-end. Big data offers an opportunity to use new data forms and emerging analytics to build processes outside-in, from the customer to corporate headquarters. This can best be accomplished when there is a team of IT and line-of-business leaders that can work cross-functionally with a focus on end-to-end processes.
This team is best led by a line-of-business leader and is guaranteed a higher level of success if it proceeds to work on the following steps.
2. Side-step religion. The term “supply chain” is fraught with issues. Some companies think of the supply chain as a limited function within the organization that focuses on logistics or inventory, while some companies think about the term as a much broader concept that encompasses end-to-end processes. Do not get entangled in arguments of supply chain as a function or an end-to-end process. Don’t argue what to call it, just get on with it!
3. Start small and iterate. Do not get caught up in the ERP-like mindset of big projects with a series of releases. Focus on small wins and learn from the use of analytics to spread to other functions. For example, the use of in-memory reporting from Qlikview and visualization from Spotfire and Tableau are being used by a number of our clients to improve data usage today to win organizational support and funding for big data initiatives. Organizations have many technologies and systems, and IT architects need to separate the decisions for analytics from the decisions being made on their systems of record implementations (ERP). ERP is only one source of data, and over time, will become a less significant contributor to the overall supply chain response.
4. Provide innovation funding. Give these cross-functional teams dollars to experiment. Allow for trial and error in the process. Some companies have had success with having departments submit requests to a cross-functional business analytics or big data team for spending on analytics and use of different data forms by cross-functional teams working on big data initiatives.
5. Consolidate business intelligence centers of excellence and master data management efforts into big data initiatives with business goals. Why? Some of the new techniques associated with advanced analytics enable data enrichment and data parsing that previously had to be hard-coded into systems. Organizations that are good at using data will win in driving big data opportunities and take advantage of these opportunities earlier. Solving business problems must be the goal. The results come by harnessing of the cross-functional efforts of knowledgeable people, working on teams to solve analytical problems.
Lora Cecere, the founder of the Supply Chain Insights research firm, is the co-author, with Charles W. Chase Jr., of Bricks Matter: The Role of Supply Chains in Building Market-Driven Differentiation (Wiley, 2012). You can read her blog, Supply Chain Shaman and follow her on Twitter @lcecere.