Big Data

Putting big data to use in the supply chain

Learn how the retail industry's smart use of big data and analytics is having positive business effects for its customers and its supply chains.

While companies continue to get their feet wet with practical applications of big data, one industry has made tremendous headway in navigating through its big data and managing it to improve competitiveness and business excellence.

I'm talking about the retail industry -- and the improvements that smarter data use and analytics have delivered not only to retail customers and customer order fulfillment, but to the supply chains behind all of these business processes that take the orders and then drive down requirements to suppliers in all corners of the world that in turn deliver a chain of finished goods to the customer.

Here are some real "difference makers":

  • Using incoming Internet of Things (IoT) sensor data from its network of service stations and convenience stores, a fuel retailer knows in near real time what its fuel product mix and consumption rates are in each area of its geographical territory. Incoming point of sale (POS) data also tells it which mixes of products (besides fuel) work best in different geographical areas so it can tailor offers accordingly. The end business result? Faster time to market with offers, and better revenue positioning and agility because offers can be instantaneously changed to respond to customer demand shifts.
  • To minimize exposure to risk in its global supply chain, a large manufacturer overlays geographical supplier locations with weather statistics for tornadoes, hurricanes, earthquakes, etc. -- and then calculates the probabilities of natural disasters occurring through a predictive analytics program. The end business result? The company now has a way to orchestrate its suppliers so that it has backup plans and failover to suppliers in other areas of the world if a key supplier gets hit by a disaster and the incident takes down production. By proactively managing its perceived risks, the company now has a way to avoid disruption to its supply chain that can endanger revenue capture and even impair customer perceptions of the company.
  • An online retailer implements real-time predictive analytics to give it visibility of which types of clothing items are selling "hottest" during peak holiday seasons. It sees that green is the most popular sweater color that is selling and quickly relays the information to its sweater supplier. Because of the strength of its analytics program, which captures and analyzes data from its website, the retailer has altered its manufacturing process with suppliers. Instead of the suppliers producing set quantities of sweaters in various colors, the suppliers now manufacture the raw sweaters without color -- and then add color based upon the near real-time customer demand instructions they receive from the retailer. The end business results? A sharp drop in the post-holiday items that need to be sold out at cheap clearance levels, and the ability of the retailer to sell more units at the highest possible price.
  • A supply chain logistics operation's largest operational expense is fuel for its vehicles. The company trains drivers in safe and sustainable driving practices, and then installs IoT sensors in all of its trucks that measure speeds that the drivers travel at, how long drivers keep vehicles in idle modes, and even braking habits. The end business result? A sizable reduction in operating expense (due to fuel savings from better driving habits) and a positive impact on the company's green sustainability initiative, since it has reduced its carbon footprint.

It is situations like these that make the retail industry a great example to others in the quest for value in big data -- and also an industry that others can borrow big data analytics practices from.

What are the critical success factors?

  • Executives in both the end business and IT were dialed in to the urgency of their situation (i.e., the need to compete fiercely and effectively in a global market, and to do it quickly).
  • Big data initiatives were targeted toward very specific business aims with measurable results that were capable of contributing to the bottom line.
  • Business processes were re-engineered because the companies realized they couldn't do "business as usual" if they were going to effectively use all of the new information that their big data was delivering.
  • These companies weren't afraid to fail in pursuit of big data initiatives. In most of these use cases, the ultimate big data goal didn't change. but companies found that they had to revise processes and systems from what they had originally thought in order to reach the goals that they set out for themselves.

About

Mary E. Shacklett is president of Transworld Data, a technology research and market development firm. Prior to founding the company, Mary was Senior Vice President of Marketing and Technology at TCCU, Inc., a financial services firm; Vice President o...

1 comments
aidemzo_adanac
aidemzo_adanac

It is finally coming to surface as a worthwhile... 'understood' practice that it necessary for retailers to offer competitive pricing and products in a timely manner.  This is the type of system I was delivering to large enterprises in 2006-2007 in the form of RFID tracking.   Such data driven supply chain systems have been in use for ages now, in the form of bar coded, asset management systems, slowly becoming more and more intricate with how data is collected and managed.  

The initial response, even from people on TR, was that it was invasive and would be used for no good.  Buy a metal wallet to stop random scanning and theft of information, wear a tinfoil hat and so on.   

There is a HUGE benefit, for the right enterprise, that trickles right down to the consumer level, but people are so reluctant to accept technology they don't fully understand, as they assume the worst and assume  insecurities in the solutions that would cause reason for public concern.  

In reality, such data is highly secured, to the best of their abilities.  It is in their best interests to secure such data as it would be invaluable to competitors too, not just data thieves with bad intentions.  

I remember people here saying, they wouldn't want someone to stand on their lawn with a $3000 RFID scanner and be able to see what type of TV they had in their living room, in order to plan a robbery.  Such conclusions are absolutely absurd though. the product itself isn't tracked that way anyway, just the packaging. Once it leaves a store, the supply chain doesn't care about the data anymore, it is irrelevant to them,   

If by some chance, a thief was to purchase a Symbol or Intermec scanner for a few grand (no, your iPhone WILL NOT do) at the VERY BEST, they can go to your garbage and scan the box to find out that it is 800234573HT482362GH.  (OH HEAVENS NO!, NOT THAT!)  OF course it would be much easier to simply walk down the alley and see what people have thrown away to identify good in their home. 

THE SUPPLY CHAIN:

Back to the supply chain then, they have taken these systems and tried just about everything, mostly with failure but with some success too.  I worked with the development teams at Symbol and Texas Instruments when these technologies were in late stages of test implementations.  The goal at that time was to have a fully automated supermarket.  Take product off the shelf, fill you shopping cart and literally push it straight out the door without even seeing a cashier.  The product is automatically scanned as you leave and money withdrawn from a registered account, VERY cool!

The retailer then has automatic stock notifications when shelves are low, this is then sent to the distribution warehouse to replenish back stock, the distribution warehouse then sends data to the manufacturer who can keep the warehouse stocked and all shipments intricately tracked from the beginning to end.  A flawless system that removes human error and keeps prices low. 

THE PROBLEMS:

While SOMEWHAT effective, the system bombed in several ways, canned drinks would work with RFID scanner (liquid and aluminum), Tetra packs (foil lined drink boxes etc.) did not scan.  The cost per item was still VERY high, with a goal of getting cost well below $25cents per tag, which has lowered now but still not in line with the fractions of  a penny barcodes cost them.

WaqlMart TRIED for force implementation from all suppliers. Considering many are smaller businesses that operates on single digit margins, they simply could not comply.  Rather than lose a high percentage of suppliers, only partial, voluntary compliance was put in place.

So it has withdrawn in widespread implementation for now, they expected costs would drop in a few years but Texas Instruments, Zebra etc just cannot provide tags cheap enough to justify many solutions...yet.

Tomorrow may be a different story but for today, while many still fear what they don't really comprehend, seeing the worst in it, such solutions are imperative for competitive prices that people simply demand these days, as quality and service are no longer a consideration when shopping, best price is the bulk of today's consumers goal, even if merely pennies.

CURRENT USES:

Include the US and Canadian military, where all materials are bar coded or RFID tagged.  This allows faster deployment of materials as individual recording and human error is taken out of the equation.  RFID means a pallet can be loaded on a plane and as it enters the plane, all items are automatically taken from the base inventory.   It helps reduce theft of items and items falling into the wrong hands.

Commercial laundry applications track clothing that leaves a retailer and follows it to the cleaning facility and back again, ensuring your clothing is not lost, again removing human error to.  

It's still young technology but will pop up more and more over time, reducing prices and making asset management more efficient for everyone.

Editor's Picks