The Shelved Shelf

By Bob Violino

What does Wal-Mart's decision to "cancel" its smart shelf test really mean?

By Mark Roberti

July 14, 2003 - Last week, CNet ran an article about Wal-Mart canceling a smart-shelf trial at a store in Brockton, Mass., a suburb of Boston (see Wal-Mart cancels 'smart shelf' trial). The article portrayed this as some kind of a setback for the use of RFID technology in stores. I got several e-mails from people concerned about this, so it's worth putting the story into perspective.

First, the fact that Wal-Mart canceled the trial was not news. Companies plan and cancel technology trials all the time. A number of proposed trails -- some public, some not -- have been canceled, delayed or moved. Which prompted me to ask retailers, in last week's column, not to hide in-store trials from consumers (see Do the Right Thing). If a trial is begun and then stopped because the technology doesn't function as expected, that's news, but that was not the case here.

To make Wal-Mart's decision seem newsworthy, CNet described the test as "one of the first and most closely watched efforts to bring controversial radio frequency identification technology to store shelves in the United States." Closely watched? The trial hadn't even begun. What were people watching?

The story goes on to say: "The benched trial was widely seen as the most aggressive step yet by a retailer to push RFID from warehouses into U.S. stores." And: "Those ambitious plans now are likely to take a backseat to proposals to upgrade warehouse operations with RFID technology, which will require fewer chips and less computational power."

Here's where we need a little perspective. Wal-Mart wasn't trying to push RFID from the warehouse to stores. RFID isn't used in Wal-Mart warehouses yet, and its focus all along has been to use the technology to improve its supply chain operations. Almost everyone's focus has been on using RFID to improve their internal operations or their supply chain. Even Gillette, which has been among the most aggressive in testing smart shelves, is not trying to push RFID into stores tomorrow. Here's what Gillette VP Dick Cantwell told me recently: "We see the smart shelf implementation being a ways down the road, but we felt that it is important to get started now."

The CNet story reported that the Brockton test was supposed to start last month. In fact, RFID Journal reported back in January that it was supposed to start later that month. I asked some Gillette executives about the test in March, and they expressed a little frustration that Wal-Mart had delayed it several times. My conclusion was that this was not a high priority for Wal-Mart. The company is very conservative and doesn't like to do any testing that might disrupt its store operations. So, it was hardly a surprise that they pulled the plug.

Wal-Mart spokesman Tom Williams was quoted in the CNet story as saying the company wants to focus on getting ready to track cases and pallets from its top 100 suppliers. Given the enormity of that task, it seems like a perfectly plausible explanation. Did media blather about privacy contribute the decision to abandon the test? I don't know for sure, but I wouldn't be surprised if it did to a small degree. I mean, if something isn't critically important and could result in bad press, why bother doing it?

The fact is, we will not see RFID tags or EPC tags on most items in grocery stores for at least a decade. As the CNet story points out, the tag price has to get down to less than a penny before that can happen. Kevin Turner told me way back in December 2001, when he was still Wal-Mart's CIO, that his company's goal has always been "to push it down to one cent a tag." This is not technologically feasible today at any volume, so it is going to require a technological breakthrough to get there.

And it's not just the cost of the tags. During her well-publicized speech last month, Wal-Mart CIO Linda Dillman showed a photo of a smart shelf. Then she showed a box of tangled wires and electrical components and said: "This is what it requires for each four-foot section in today's world." Her point was that smart shelf technology is not ready to be deployed in stores yet.

The CNet suggestion that Wal-Mart has given up on an ambitious plan to push RFID technology into stores is flat out wrong. Clearly, that has never been Wal-Mart's near-term goal. On the other hand, I do not believe that Wal-Mart has given up on the long-term vision of using smart shelves to reduce out of stocks. My guess is that when it has the technology working well in the supply chain, Wal-Mart will experiment with smart shelves at its facility in Rogers, Arkansas. And when the industry has agreed on privacy policies that consumers are comfortable with and when smart shelf technology improves to the point where it can be tested without affecting in-store operations, Wal-Mart will do serious trials.

Many publications that picked up the Brockton trial "news" made a big deal of the privacy issue and the impact it may have had on Wal-Mart's decision to scrap the test. But here's the most interesting part of the story, which everyone seems to be missing. Privacy advocates are concerned that companies plan to use RFID technology to track individual consumers. Yet, the world's largest retailer just made it clear that it not only has no interest in tracking individual customers; at this point, it's not even terribly interested in tracking individual products.

That should be reassuring to people concerned about privacy. Everyone else should be reassured by Wal-Mart's commitment to using RFID technology in the supply chain, even if tags aren't cheap enough to put on individual items. Maybe we should all stop talking about tags on products in stores and focus on the supply, which is what it's all about right now anyway.

Mark Roberti is the Editor of RFID Journal. If you would like to comment on this article, send e-mail to

RFID Journal Home