Monday, March 19, 2012

Big Data in Supply Chains

imageSuddenly, the big data is everywhere. Big data is huge (pun intended). The buzz around big data itself is pretty big. How big is big data after all? IDC says that in 2010, the size of digital universe cracked the zettabyte barrier. How big is a zettabyte: That would be a giga-giga-kilo – that is 21 zeroes after 1 or 1,000,000,000,000,000,000,000.

This data consists of everything: Facebook, Twitter, YouTube, Emails, Web Commerce, Blogs and everything in-between. Almost every click becomes someone’s data that they are collecting to harness some useful information, most of it in the hope of understanding and then managing consumer behavior for commercial and business purposes. 
According to McKinsey Global Institute (MGI), US Library of Congress had 235 terabytes of data by April, 2011. It further states that 15 out of 17 sectors in the United States have more data stored per company than the US Library of Congress. What is the commercial potential of this data? According to MGI, there is a 60% potential increase in retailers’ operating margins possible with big data.
MGI identifies several functional areas where big-data driven efficiencies are possible. No surprises among them: Marketing, Merchandising, Operations, Supply Chain, and new business models.
How will supply chains become part of the big data revolution? How ill it affect the supply chain solutions and technology? How will it provide another opportunity for the supply chain pioneers to create competitive advantage? Here are some starters:


It was a big hype, then it was not, then it was! No matter how you look at it, RFID has the biggest potential to change how supply chains work. While cost of RFID chips has been declining, the biggest reason for low adoption rates has simply been the inability of the companies to leverage the large streams of data produced. What can SKU-level RFID information provide?
  • Real-time inventory of of any item in any location: This is a big deal, a really big deal. Some retailers may have this information currently, but it comes at a high cost and is usually suspect due to human intervention. But automated shelves with RFID readers can automate the whole affair, reduce the errors, and even pinpoint the location of items in a store or warehouse when they are misplaced!
  • Automated replenishment signals can follow, allowing inventory analysts in the stores to support customers on the floor.
  • Automated receiving and verification of items and quantities received at stores and warehouse: This can be a big deal simply on labor savings, but it also automates the settlement process.
  • Automated validation of fulfilled orders: Are you shipping what the customer ordered? How about the label on the box: Does it have what it says it does? Correctly filled orders enhance customer satisfaction and perfect order metrics.
  • Automated checkouts: Think beyond the self-checkout counters – how about simply eliminating them altogether?
  • Maintaining the cold-chain integrity: Using RFID chips can provide verified temperature controlled merchandise throughout the transit.  

GPS-based location services:

While the GPS-based shipment management is entering mainstream, most retailers have yet to scratch the surface when it comes to leveraging this technology to dynamically manage supply-side of the supply chain.
  • Most retailers have sizeable inventory in-transit at any given point of time. Most often, this inventory is simply headed for a warehouse and does not get into the fulfillment calculations unless physically received. But having real-time visibility of this inventory can work wonders. The ability to sense real-time demand signals (from RFID-based automated checkouts, may be) and combine them with real-time inventory locations including inventory in-transit can make this inventory “productive”. Real-time location sensing can make sure that this can be dynamically deployed for best supply-demand match, and reduce the fulfillment lead-time and inventory levels required without affecting service the customers.
  • Combining the RFID for cold-chain perishable goods and real-time GPS location, efficiencies can be gained to reduce the goods damaged due to temperature variations and expiration dates.

POS data:

POS data is probably the one part of “big-data” supply chain processes that have been leveraged to an extent by some pioneers. P&G and Wal-mart sharing the POS data to collaborate better is well known. What else can a POS data stream do for you?
  • POS data can provide a real-time demand signal with price information. As demand velocities change, it allows for intelligent inventory deployments to optimize the inventory in the system, but the price information embedded in the demand signal can actually help optimize the profitability when the supplies are scarce.
  • Early trends detected for seasonal goods can help better manage the open orders and potentially help the top-line when demand goes up and reduce potential clearance losses when it goes down.
  • Price optimization can be fine-tuned with real-time POS data to optimize the profitability.
  • Demand collaboration with the suppliers can not only create more trusting relationships, but actually drive buy-side efficiencies. Creative solutions will include committed percentage of business, vendor-managed inventories in the warehouses, and replenishments in the store.
Related Articles:
Other Resources:
Want to know more about supply chain processes and supply chain strategy? Check out my books on Supply Chain Management at Amazon.
© Vivek Sehgal, 2012, All Rights Reserved.

1 comment:

  1. Nice blog. Splunk is a solution that can bring together POS, RFID, and GPS data, correlate it and run automated statistical analysis in real time at the petabyte scale.

    I'd be interested to know what risks might be mitigated or what risk related questions you could ask of your data if you could get all three of these data sources into a big data solution that could run analytics on the data and represent it visually.


    Mark Seward