Evan Schuman's StorefrontBacktalk

Franz Dill spent 30 years in various technology management roles at Procter & Gamble, the world’s largest consumer goods manufacturer. He retired from P&G and joined the consulting world in 2007. This is a StorefrontBacktalk GuestView column.

When Wal-Mart this week confirmed its limited launch of item-level RFID tags, it shouldn’t have been news to many in retail. That’s what Wal-Mart has said it wanted to do for literally 10 years. The question shouldn’t be “Why is Wal-Mart doing this?” as much as “What’s different this time? Why is this finally happening now?”

It’s not a technology issue. It’s a mindset issue, in two ways. First, Wal-Mart’s “let’s start at the pallet level and learn RFID there” approach was an understandable albeit wrong move, given the retailer’s data and analytical needs and goals. Second, the need for the kind of data that only item-level can deliver is now much more apparent and, perhaps more importantly, understood better at the management level.

The way to start is small, using item-level RFID–but only on select items–and then using RFID everywhere else in the supply chain. So how did Wal-Mart go so wrong for so long?

The company’s initial pallet move was done primarily for one reason: It was easy. But it was a mistake to start at the pallet level, because Wal-Mart couldn’t measure the actual benefits it needed to understand. Meaningful measurement occurs out on the shelf, when the consumer makes a choice.

The pallet/case approach was also a mistake in the broad view. It had value for testing but not for the ultimate understanding of “What am I attempting to achieve? Which things are most valuable for me to do?” To expect to get information about things like out-of-stocks and what’s on a display from pallet tagging is unlikely. And yet, I do think Wal-Mart expected to get out-of-stock information directly through its pallet-level tagging efforts.

To understand Wal-Mart’s approach, it’s important to remember RFID’s history. Back in 2000, when I managed the retail innovation center at Procter & Gamble, we positioned RFID tags to soon replace barcodes for retail item-level identification. Indeed, I gave presentations to many hundreds of innovation center visitors saying the switch from barcodes to RFID would occur in 3 to 5 years. So why didn’t it happen at that time, when several large companies were willing to spend lots of money to push the tags?

Early indications suggested there would be benefits from making the entire extent of the supply chain more transparent to those who paid for it: retailers, manufacturers and, ultimately, consumers. But imagine how complex it would be to tag something and have it be readily scan-able in the entire supply chain: at the checkout, in the consumer’s shopping cart and by the shelves themselves to verify inventory. The tags would also be used at retailers’ doors to check if an item had been paid for. In addition, the largest players would benefit first and most. Clearly, the complexity of these changes would add cost.

And yet costs would be saved through transparency within the supply chain, in improved store stocking and even in the homes of consumers, where they could remotely check their personal inventory of goods. And with the emergence of devices like smartphones, consumers would be readily able to participate in this improvement in retail..