Cracking the Code

MICHAEL GARRY

In the beginning, there was a 10-pack of Wrigley's chewing gum sporting a UPC bar code. It was passed over an NCR scanner at a Marsh supermarket in Troy, Ohio. The laser light from the scanner measured the distance between vertical bars of the bar code, thereby identifying the product. Its price was then looked up in the system and displayed.

It was June 26, 1974. Bar code scanning was born, and with it the modern supermarket industry.

The impact of the bar code on food retailing has been such that one can easily divide the history of the industry into before the bar code (B.B.C) and after the bar code (A.B.C). From checking out customers and taking inventory on the shelf, to processing shipments from the warehouse and analyzing store performance, every aspect of the food retailing business is more efficient and less costly because of the bar code and the data it makes possible.

“We've seen the bar code evolve into much more than a simple solution to check out groceries and remove price markings from supermarket shelves,” said Ed Crenshaw, chief executive officer of Publix Super Markets and a member of the board of GS1 US, which manages the UPC in this country. “It touches virtually every area of our business. Because inventories are more accurate, labor is better deployed and marketing dollars are spent more efficiently, our customers get better value. It's been a win/win for everyone.”

The last time an estimate was made of the financial impact of the bar code — in 1999, on the 25th anniversary of the first scan at Marsh — PricewaterhouseCoopers, New York, calculated that it saves the U.S. grocery sector alone $18 billion annually. Ten years later, it is used by more than 200,000 manufacturers in the U.S. and has been adopted internationally by more than 100 countries.

In 2009, 35 years into the bar code era, it is worth assessing the impact it continues to have. But it is also worth asking: Has the food industry leveraged the full potential of bar code technology?

A PRECIOUS ASSET

A good example of what the bar code has made possible in 2009 is the technology infrastructure assembled by Haggen, Bellingham, Wash., which operates 32 supermarkets under the Haggen Food & Pharmacy and Top Food & Drug banners.

Over the past two years, Haggen has implemented an enterprise data warehouse system from Teradata, Dayton, Ohio, which gives the retailer the ability to store and use detailed POS data accumulated through the scanning process. At the same time, Haggen has invested in business intelligence software from Manthan Systems, Schaumburg, Ill., which allows in-depth analysis of POS scan data residing in the data warehouse to support category and inventory management.

Haggen's data warehouse allows the chain to keep data “at the most granular level,” said Harrison Lewis, the chain's chief information officer. The Manthan business intelligence system added a “rich retail data model and a rich set of KPIs [key performance indicators],” such as sales per hour, promotion effectiveness and inventory turns. The chain is now able to rapidly build applications on top of the data warehouse at a very low cost, such as order management, transportation cube optimization and loss prevention.

The chain further enhanced its technology infrastructure with a system from GoldenGate Software, San Francisco, that allows POS transaction (t-log) data to be accessed in near-real-time. And its latest innovation is a system from Accelitec, Bellingham, Wash., that drives loyalty via a number of unorthodox applications.

All of these systems are based on POS scan data, Lewis noted. “I look at POS data as one of the most precious and valuable data assets a retailer owns,” he said. “It's critical for us. It's front and center.”

Because of the importance of the data — and the need to act on it as soon as possible — Haggen invested in the GoldenGate system, which also serves as a defense against a server crash or other snafus blocking access to the data. “I said we need to guarantee we get the data, improve quality and completeness, and eliminate latency,” said Lewis. The system reads a t-log in sub-seconds and sends the information in XML format to the operational data store (ODS) at Haggen's headquarters, where it is moved into the data warehouse every 15 minutes.

As a result of this speedy access to POS data, Haggen is able to conduct what Lewis called “intra-day” activities within a single business day. For example, if a promotion is started and chain executives want to know how it is progressing, they can be informed within 15 minutes rather than the next day. “In most cases the next day is OK, but if we've got one store or area doing a better job than others, that could be a call to action. What are they doing differently? We could convey that to the other stores and improve overall performance.”

If a slow-moving product is being promoted, he noted, it may be susceptible to being out-of-stock because of limited holding capacity. “We are able to know about a potential out-of-stock and take action to refill the shelf or create an endcap,” Lewis said. “We don't have to wait until tomorrow. It prevents lost sales.”

Lewis described the evolution of how POS information is obtained. “First it would be once per quarter from IRI, then once a week, then once a day,” he said. “Now we get it intra-day, which is becoming the norm. It shows the power of having the information in near-real-time.”

Haggen's aggressive use of POS data also demonstrates that the data can be leveraged in this manner by a relatively small food retailer. More commonly, giant food retailers like Wal-Mart, Kroger, Safeway and Tesco have been the leading practitioners of what consultant Thomas Murphy has called “Molecular Management.” This style of management uses real-time or near-real-time data to support marketing, inventory, pricing and other decisions, said Murphy, president of Peak Tech Consulting, Colorado Springs.

Tesco has famously been able to gain market share in its home market in the United Kingdom by partnering with loyalty card guru Dunnhumby to mine the purchases made by card holders and then market specifically to their tastes. Kroger followed suit in the U.S. by forming its own partnership with Dunnhumby while Safeway used POS data from its loyalty card program as the basis for implementing its lifestyle store format.

No less than these retail behemoths, 32-store Haggen “has transformed its business from a traditional grocer to an information-driven enterprise,” Lewis said during a presentation at Teradata's users conference last month. He later told SN, “Our needs are as great as, if not greater than, a large retailer's. We have fewer stores but our business is more complex because of our assortment, localization and emphasis on freshness, service and prepared meals.”

PROMISE FULFILLED?

While chains like Haggen are investing in technology that exploits POS data, has the food retailing industry as a whole lived up to the promise of bar code technology? In its 1999 report, PricewaterhouseCoopers noted that even though the grocery industry had exceeded what was originally expected in 1974, it had still fallen far short of what was possible, despite such initiatives as Efficient Consumer Response. Has that gap been bridged in the ensuing decade?

One of the applications that feed off POS data is computer-assisted ordering (CAO), also called computer-generated ordering (CGO). With a steady flow of POS data, coupled with carefully tracked inventory, stores could allow computers to forecast replenishment requirements and create store orders for warehouse deliveries. In the ultimate scenario, that data would filter up and be used to generate warehouse orders for suppliers. Product could be manufactured on a “just-in-time” fashion, making one product for each one sold.

One retailer that has embraced CGO is Price Chopper Supermarkets, Schenectady, N.Y., which applies its system to not only warehouse-delivered Center Store products but also perishables and direct-store-delivery items. For Center Store products, the chain has reduced its out-of-stock levels to 2%, from as much as 4.5%, said Mark Chandler, Price Chopper's vice president, supply chain.

“Obviously, POS data is extremely important in our CGO/CAO efforts,” said Chandler. “We use POS data to identify prior item movement and that prior movement data is used to forecast future movement.” To get an accurate forecast, he added, the POS data is paired with upcoming promotional activity, price points, potential coupons, in-store display plans, among other factors.

But CGO has been a slow go for other retailers. Among the top 20 food retailers, “only a handful are doing anything sophisticated with forecasting and automated replenishment,” said Mike Griswold, vice president, research, AMR Research, Boston. The biggest challenges to adoption, he added, have been “selling the business case for store-level inventory management as well the significant change management implications.”

More than their internal application of POS data, the PricewaterhouseCoopers report was especially critical of food retailers' failure to share data with their trading partners. By contrast, the ability of alternative formats to share data with their suppliers — such as Wal-Mart through its RetailLink portal — was considered a key reason they were able to create a cost advantage over traditional grocers and make inroads into the food distribution business.

In the past few years, food retailers have stepped up efforts to share POS and other data with suppliers. A primary example is the “New Ways of Working Together” initiative launched by Wegmans Food Markets, Rochester, N.Y., with suppliers like Coca-Cola, J.M. Smucker and Kraft Foods, in order to foster more productive trading partnerships.

In one example, Wegmans provided Coke with historical and daily POS data that was used to calculate a forecast of future sales. “POS information was the foundation of the project,” said Kraig Adams, director of collaborative industry development for Coke, at last year's U Connect conference.

New Ways of Working Together “has been a long time coming,” said Griswold, but “might actually have some stickiness.” But overall, he added, the industry's progress with data sharing has been slow, and it “still has a long way to go.” In addition, although the focus of collaboration has been around grocery products, “the real value will be getting to perishables where everyone is looking to create differentiation,” he said.

But a white paper, “Retailer-Direct Data Report,” recently published by the Grocery Manufacturers Association, suggested that significant progress has been made with respect to retailers sharing POS data with suppliers. “Today, most U.S. mass merchandisers and grocery retailers with more than $5 billion in annual sales are already sharing data directly with their suppliers free of charge,” the white paper reported.

The paper identified more than 30 business benefits for manufacturers who use retailer-direct data, including reduced out-of-stocks, lower inventory and advanced price and promotion optimizations. “Investments made in this practice will pay dividends for many years as the data fields get bigger, come faster, have more depth and detail, and as retailers turn over more decisions and actions to the oversight of suppliers,” the paper said

Another manifestation of greater collaboration made possible by the bar code is the growth of data synchronization. In synchronizing their product data via the Global Data Synchronization Network, a number retailers and manufacturers use a product's global trade identification number (GTIN), which is based on its UPC code.

“Before the bar code, each manufacturer gave internal codes to their products,” said Donal Mac Daid, vice president of product marketing, Aldata Solution, Atlanta. “Suppliers and retailers would need to figure out ways to transform the manufacturer codes into their own language. That is no longer a problem.”

THE FUTURE

Where is the bar code headed? The biggest change in bar coding since 1974 is the GS1 DataBar, which is smaller than the UPC but holds more data. One version of the DataBar is being applied to loose produce, sharing space on the sticker with the traditional PLU number, while another version is found on coupons along with the UPC-A code.

Retailers are expected to adjust their scanning systems and POS software to be able to scan and process the produce DataBar by Jan. 1, 2010, the so-called sunrise date. The sunrise date for coupons — when manufacturers can begin using only the DataBar on coupons — was originally the same, but was postponed one year by the Joint Industry Coupon Committee when it determined that many retailers' POS systems wouldn't be ready by 2010.

Not that retailers have to wait. “Despite the deferred implementation position of the committee, the JICC encourages retailers to proceed with implementation of the GS1 DataBar as soon as they are ready and not wait until the 2011 date,” said Stephen Sibert, GMA's senior vice president of industry affairs.

One of NCR's retail customers has been scanning the DataBar on coupons for the past month and has benefited from the electronic verification of expiration dates, according to John Wilson, senior product manager, bi-optic scanners, NCR. “There was apparently a lot of redemption of expired coupons under the old system,” he said, declining to name the customer. Another retailer has been scanning the DataBar on produce since last year, he added.

While the DataBar represents an advance in bar code technology, many observers believe that bar codes will ultimately be replaced at the item level by RFID tags. Those tags, which respond to radio waves emitted by a nearby reader, are able to identify each specific item (this particular can of Coke, not just a can of Coke), and do not require a clear line of sight to the reader as bar codes do to the scanner.

RFID tags containing data in the form of an electronic product code (EPC) have been successfully used on pallets and cases (largely by Wal-Mart). But due to their cost — about 5 cents apiece at a minimum, compared to about half a cent for a bar code — they won't be used in grocery at the item level “for at least 10 years,” said Joe Zenobio, chief product and solutions officer, GS1 US, Lawrenceville, N.J. “They need to be below a penny,” he added.