When Wegmans Food Markets began synchronizing product data with its suppliers about six years ago, the Rochester, N.Y.-based chain expected “a seamless flow of information, and the elimination of errors and disruptions,” said Kristin Andersen, project manager for Wegmans, in a session at the U Connect conference in June. “We were really excited.”
Data synchronization of product attributes between retailers and suppliers has been heralded as a way for trading partners to use the same product information in their respective organizations, thereby preventing the kind of invoicing errors and other snafus that often bog down communications and commerce. And that can be achieved — as long as the data being synchronized is accurate, i.e., reflects the true physical dimensions and weight of each product and its shipping container. Otherwise, the whole exercise is for naught.
Inaccurate product data has proved counterproductive to the efforts of Wegmans and other retailers pursuing data synchronization programs. “We found we were populating our database with really bad information,” said Andersen. “Our business users said we had bad data and weren't using it.”
Since then, Wegmans has spearheaded a pioneering campaign to clean up its data by, in effect, having its manufacturers clean up the data they were sending the chain. Wegmans thereby sparked industrywide interest among retailers and manufacturers in addressing the data accuracy issue.
Still, data quality remains an issue that many companies prefer to ignore. “It's the inconvenient truth that no one likes to talk about,” said Steve Vazzano, director of data quality for GS1 US, the standards organization based in Lawrenceville, N.J.
Though suppliers are thought to be responsible for the accuracy of their data, retailers like Wegmans and wholesalers like Wakefern Food Corp. often find themselves the final arbiter of whether data passes muster and can be safely used in their companies. That typically involves auditing the physical dimensions and weight of products and checking whether that data matches the information on hand. To help retailers and manufacturers with that onerous task, new third-party certification services are beginning to emerge (see story, this page).
In 2006, Wegmans found that only 30% of its products had accurate product data for attributes like size and weight. Three years later, following extensive collaboration with a wide array of trading partners, Wegmans has its accuracy average up to 79%. “This is about working with trading partners, developing strong and trusting collaborative relationships,” said Andersen. “And by working on data, we could begin working on other things important to our business.”
The value of accurate data goes beyond just data synchronization. Once the data resides in Wegmans' master database, it is then used by all of the retailer's operating departments, including transportation, warehousing, shelf planning, procurement and its website. “Data quality allows us to use the information to run our business,” said Andersen.
Some product data is so obviously off the mark — a 7-foot bottle of shampoo or a 53-pound bag of candy — that it's easy to spot and correct, noted Andersen. More dangerous are the small inaccuracies that insidiously accumulate. “An inch here, a pound there, it all adds up and has an impact.”
For example, by correcting a 1-pound discrepancy on just a single item, Wegmans can take six trucks off the road in the course of a year, saving an estimated $10,000 in fuel and freight charges.
Vazzano of GS1 US pointed out that the savings stemming from data quality can reach six and seven figures, though companies don't like to acknowledge those savings publicly.
At the store level, accurate dimensions ensure that planograms are properly designed and that each product gets its share of space. Net content is another attribute that needs to be correct, Anderson noted; if it isn't accurately stated on shelf labels, retailers may be subject to fines.
Andersen shared some best-practice tips for achieving accurate data. For one, companies need executive-level support and at least one employee who can oversee data quality. Companies also need a master data management strategy that ensures the maintenance of “one version of the truth.” In addition, there should be procedures for maintaining accurate data and ensuring that a retailer is working with data for finished goods, not products that are still “on spec.”
Hershey, Hershey, Pa., and Kraft Foods, Northfield, Ill., are two suppliers that have boosted their data accuracy numbers as measured by Wegmans over the past year, from 83% to 100% in Hershey's case and from 67% to 97% in Kraft's.
Kraft has been working over the past three years to establish internal control practices in order to ensure ongoing data quality. “Collaborating with Wegmans on data accuracy has been a very rewarding experience,” Karen Spooner, global standards and GDS associate director for Kraft Foods, stated at U Connect. “It has helped us improve our business processes, which have led to efficiencies in transportation and warehousing for both Kraft and our [retail] customers.”
To determine where data integrity could be enhanced, Kraft leveraged the help of a third-party expert, Gladson Interactive Services, Lisle, Ill., to conduct an internal data audit. “We collected 150 attributes per SKU and took photos of every consumer unit, case and inner-pack for all of our U.S products with available inventory,” said Spooner.
Spooner pointed out that a data accuracy program requires top-management support and a single master database that feeds all company processes. “If you rely on a master database, you will have better controls, which will lead to greater accuracy,” she said.
Also important are standard data procedures that everyone at the company follows, making education and training key to a successful data quality program. In addition, Kraft developed a data certification program that continuously monitors new and existing items, and provides the certification status of a product at any stage of its life cycle.
In setting up its data accuracy program, Kraft made use of the Data Quality Framework assessment tool, which was developed by the CPG industry with the support of GS1, Brussels, the global standards organization. The framework includes an assessment procedure to validate the existence and effectiveness of key data management business processes, as well as an inspection procedure to physically validate product attributes. “It gives you an idea of the controls you should put in place,” said Spooner. “It's a great guideline.”
Maintaining data quality is hard work, and there is no software package or plug-and-play solution to make it easier, said Spooner. As a manufacturer, Kraft found it helpful to be monitored by a retailer such as Wegmans. “You don't know where your opportunities reside until you understand the impact from a retailer's eye,” she said. Data managers also need to collaborate with internal departments at their own company, she added.
“Data quality is not just for data synchronization,” Spooner said. “CPGs should view it as a business driver. Master data should be viewed as an asset.”
When Hershey was confronted with the initial audit results from Wegmans, “We didn't realize we had bad data,” said Rob Hoffman, Hershey's program manager. “We said, ‘It's our data, why would it be wrong?’”
The key for Hershey, Hoffman said, was to see the proof of the inaccurate data from Wegmans in the course of “top-to-top meetings” between Hershey and Wegmans executives. “That was the tipping point for us, because we need to give our [retail] customers what they want.”
Hershey began its data accuracy overhaul in January 2007 and completed it 17 months later. “We did a clean sweep of all items, which ended up including 1,995 GTINs,” said Hoffman. The company employed an automated process for measuring products and capturing weight and dimensional data that included a scanner, calipers, a scale and a laptop computer.
In educating its internal teams on the intricacies of data accuracy, Hershey recruited package measurement rules experts from GS1. “We had a full day at a warehouse and then a road show at our company for the marketing, logistics and packaging departments,” Hoffman said. Different Hershey warehouses now share data accuracy best practices.
In addition to cleaning up its data, Hershey needed to implement a process that would quickly incorporate accurate data for new items. “We need to capture data before the new items hit distribution,” Hoffman said. Since November 2007, the company has used the data capture process for 1,400 new items.
The new item data collection process at Hershey involves the submission of a new item information form within 24 hours of the emergence of new inventory, followed by the updating of the master database with accurate information. If the form is not received, the appropriate facility is contacted and given a timeline to submit it. If the form is still not received, an executive request is made.
The new item form has eliminated the risk that Hershey will ship new items to retailers with “estimated data” from products still on spec, said Hoffman.
He enumerated several other benefits of accurate data for Hershey. One is that “[retail] customers now trust our data,” he said, eliminating the need to send product samples in advance for data verification, a costly process. And like Wegmans and Kraft, Hershey has realized considerable transportation savings from accurate data.
Accurate data also delivered a host of benefits at the retail side, including more efficient use of transportation capacity, fewer invoice deductions, fewer out-of-stocks, better use of shelf space and more time for sales planning. “We're no longer wasting a lot of time on whose data is accurate,” Hoffman said.
While Wegmans has probably been the most proactive food retailer in working on data accuracy both internally and with suppliers, other retailers and wholesalers are beginning to follow suit. For example, Wakefern, a cooperative wholesaler based in Keasbey, N.J., has made a commitment corporately and to its more than 40 grocery members to “improve the foundational information of our enterprise,” said Michael Durning, Wakefern's manager of data integrity.
“The objective came from both sides,” he said at U Connect, “from the retailers who said we need to improve the quality of information at our stores and from within our corporate organization.”
Like Wegmans, Wakefern uses a master item database designed to be the sole source of accurate data for its internal applications and processes, as well as its members and wholesale customers.
At the store level, accurate data, such as net content on shelf labels, protects Wakefern from fines for errors that are issued by the Department of Weights and Measures in New Jersey, where many of its member stores are located.
At the supply chain level, Wakefern calculated that if the weight shipped from Wakefern facilities is overstated by just 1% for one year, the result is an additional 1,507 loads that would need to be shipped, or 30 additional trucks per week, not to mention fines for overweight trucks, Durning said.
Wakefern has developed a three-pronged approach to keeping its data accurate. One is to engage in data synchronization “with a focus on receiving quality data from our suppliers,” said Durning. The wholesaler also recently completed the development of a new item portal intended to eliminate paper forms and manual keypunching; Wakefern just began to pilot the portal with a small number of grocery suppliers. And Wakefern is creating a centralized department focused on “quality and control of our master item data,” he said; the department so far supports five product divisions.
In its data synchronization effort, which uses 1SYNC as a data pool, synchronized data is integrated into Wakefern's master item database. All item attributes are compared with Wakefern's own information; any discrepancies are reviewed, with Wakefern reserving the right to challenge suppliers via an email notification. Wakefern has established a website for data synchronization at www.shoprite.com/cnt/Wakefern_GDS.html.
Last September, Wakefern send its first letter to suppliers requesting them to support data synchronization via the Global Data Synchronization Network. As of May 1, the cooperative had established 285 connections with trading partners and synchronized 12,000 items representing 24% of slotted warehouse items.
In June, Wakefern sent a second letter to suppliers, asking them to engage with a data pool, clean their data and begin publishing data to Wakefern by Aug. 1.
Wakefern's New Item Portal is a Web-based application integrated with data synchronization and available to suppliers and brokers with secured access. Suppliers are able to track where new items are in the rollout process. By eliminating item forms, the portal is designed to increase efficiency and speed new item introductions, said Durning.
Wakefern's centralized data integrity department establishes data standards for the organization, acquires knowledge of how item data is used at Wakefern and by member stores, and maintains control of the Master Item Database for the life of an item. The department also handles all new item introductions and performs validation checks and audits for all new items.
For existing items, the department reviews all data changes received through data synchronization, performs audits and works with suppliers to resolve data integrity issues.
The data integrity department serves as Wakefern's single point of contact for stores and suppliers, “so we can give a consistent response,” said Durning.
Durning urged suppliers to support data synchronization and adhere to industry guidelines such as GTIN allocation rules and product measurement rules. Suppliers “own item data and [they] are responsible for its accuracy,” he said.