Wal-Mart's Item-Level Strategy: Better That Tags Should Be Thrown Out Than Dealt With

When Wal-Mart this week confirmed it has been quietly testing item-level RFID in two Arkansas stores for several months—along with plans to "incrementally roll out [item-level RFID] throughout the chain"—it raised quite a few eyebrows because of the way it's being done. The company is initially only tagging denim jeans, socks and underwear (let's try and ignore the fact that a radio transmitter inside a guy's boxers is nothing shy of creepy), and it's leaving the tags active until customers opt to throw them away.

The reality is that Wal-Mart's gradual deployment makes a lot of sense. The media-repeated cries of privacy invasion are simply silly, based on ludicrously unrealistic assumptions of how easy this data would be to access, assuming anyone had any reason to even try. The most interesting part of the rollout, though, is the tag disposal.

The way Wal-Mart has set up its item-level RFID EPC system, the tags will be attached (not integrated inside, but attached) by the manufacturers. The chain will track the tags throughout its supply chain, all the way to the shelf and the cart. But it won't be scanning for them at the POS, so the path ends somewhere shy of checkout.

That's where things get interesting. The tags are supposed to be easily removable by consumers. (Please don't get me started on things that are supposed to be easily removable by consumers. And, yes, I'm looking at you, Costco. When someone can make a living selling devices to let consumers access products they've already paid for, it's time for you to consider whether you've perhaps overdone your packaging a bit. But I digress. Back to Wal-Mart item-level.)

So why not simply have the Wal-Mart associates remove the tags at POS and then throw the used tags in a box so they can be reused later? Staples made quite a case study a few years ago with its reusable active RFID tags.

This approach would theoretically accomplish lots of good things: It would save the consumer the bother of removing the tags; eliminate these baseless privacy attacks that envision cyberthieves scanning people's garbage cans to learn about their jeans; and save a few dollars because the tags could indeed be wiped and reused.

Even if the tags are being purchased for a penny or two, at Wal-Mart volume levels, that has to add up. Besides, they are literally going to be thrown away. These are sophisticated pieces of electronics that the chain needs anyway. Why let that happen?

There is a practical reason for letting the tags leave the store. If Wal-Mart's volume purchases do indeed get the price down to one cent per tag--which is far from out of the question--you have a simple ROI issue. How long will it take an associate to remove each tag and place it in a receptacle? How many of the tags will be damaged by that act of removal? How much will it slow down checkout lanes? If a brand-new tag can be purchased for a penny, it wouldn't take especially large answers to any of those three questions to make it not worthwhile.

Let's take a quick look, though, at that privacy issue. Like everything else, it's a lot more complex and nuanced than we want it to be.Let's take a quick look, though, at that privacy issue. Like everything else, it's a lot more complex and nuanced than we want it to be. At a technical level, the claims that people can be tracked when they leave the store--and that bands of nosey marketers and cyberthieves will be war-driving garbage cans--simply doesn't pass the laugh test.

First, the scan ranges are simply far too short for those scenarios to work. Second, too many things could interfere with a clean scan even if it was feasible, such as used aluminum foil, soup cans or liquids. Third, the data provided is simply not worth the effort.

And fourth, is it even remotely practical to drive around on the off chance of stumbling upon a consumer who was a Wal-Mart customer plus shopped at one of the few stores in the pilot plus happened to have purchased one of the few items in the pilot plus happened to have thrown out the tag in that garbage can (as opposed to throwing it out anywhere else or simply left it attached to the garment in a drawer)?

Another privacy concern that has been expressed has been that the signals being broadcast from item-level tags will be privacy-dangerous when combined with signals being sent from chipped driver's licenses, contactless payment and even passports.

This is another one of these concerns that sounds frightening on paper but doesn't play in the realworld. The chains that can't associate a customer's online purchases with their in-store activity, nor remember a conversation they had with a call center rep nor store notes from an associate they spoke with for 50 minutes in-store (note that all of these transactions happened with employees with the same retail chain), these systems are suddenly going to seamlessly integrate data from various wireless IDs that might briefly pass by wireless readers?

You don't even need to get to the backlash against them trying nor the costs involved in monitoring and analyzing those intercepted transmission nor the extremely little usable data likely to be discovered nor the fact that there are so many easier ways to gather usable CRM data of that kind. No, you can simply focus on the realworld logistical issues and conclude that this simply not at all likely to work any time soon.

That all said, reality is not necessarily the determining factor. Even though the reality is that privacy is not in any way challenged by Wal-Mart's RFID item-level effort, if enough consumers believe it is, it will become a problem. Paradoxically, the more Wal-Mart tries to convince consumers that there's no privacy issue (regardless of how true it might be), the more consumers might believe the opposite.

That's why removing the tags at POS might make very good sense for Wal-Mart--and other chains considering similar efforts--even if it doesn't necessarily make ideal spreadsheet sense. Unlike credit card thefts, which are blunted by zero-liability programs, a widespread belief that privacy is threatened more at Wal-Mart than at Target or Sears could actually start impacting revenue.

The simple act of removing the tags at checkout could make a lot of sense for true reasons--recycling the tags--and bogus ones--privacy. An approach that makes sense for both reality and make-believe? What a concept.

In a wonderful StorefrontBacktalk GuestView column this week, Franz Dill--who spent decades working for Procter & Gamble's technology efforts, including RFID--makes a powerful case for why Wal-Mart's RFID attempt will work much better this time than it did a decade ago.

But the key difference is money. Yes, time has allowed for Wal-Mart to better understand RFID and to tweak scanning techniques. That doesn't explain this change, though. Now, Wal-Mart officials concede, they are subsidizing supplier efforts, reportedly to a much greater degree than before.

Wal-Mart's early efforts were marked by the world's largest retailer dictating to suppliers that they would support EPC tagging and that they would do it immediately and for free.

"What's in it for us?" suppliers asked, in unison. "Simple," Wal-Mart replied. "Revenue enhancement." How so? "Right now, you make $X million from Wal-Mart sales. If you don't do this, you won't make any. Capisce?"

The response to Wal-Mart's subtle messaging proved there was indeed a point where Wal-Mart suppliers grew backbones. They resisted, and the initiative failed.

Today, however, many years and even more pallets later, Wal-Mart's game with suppliers of "Tag, you're it" has become "Tag, we're going to try."