Rather than wait for an industry to bless a standard for RFID or a data requirement for each chip, the Dutch retailer BGN launched a major strategic initiative to build ‘the book retail store of the future’ using self-designed RFID tools and technology. The resulting implementation has been the talk of the RFID industry and why not since this relatively small retailer operating in a relatively small industry has done what the mighty Walmart has been unable to do. From the outset, success for BGN was more assured because the company embarked on this effort for its own business reasons rather than having to implement it due to some directive from a third party.
The Dutch experience was presented at a panel meeting on Saturday at BookExpo to a disappointingly small group. RFID is in the process of transforming the BGN business and while publishers and retailers from the US have visited their operations to see the implementation in detail there doesn’t appear to be immediate impetus in the US to launch an RFID initiative. The Dutch example is notable for several reasons: Firstly, the company approached the initiative from a complete supply-chain view and they recognized that they needed to involve other parties in the planning and design of the initiative. Secondly, the company was first to implement the solution, and as a result, key software and hardware vendors were more willing to be flexible to support the implementation since it was in their interests to succeed. Thirdly, the company used experiences gained by others-such as Borders’ use of kiosks, Gartner’s research on store design and Metro of Germany’s test store.
The retail (not just publishing) industry suggested it couldn’t be done, arguing that privacy, tag costs and product breadth were issues too difficult to overcome. While BGN understood and addressed these concerns, Matthijs vd Lely, CEO of BGN, commented while on the Panel that, even prior to implementation, some executives remained skeptical.
The results have been more than impressive. At the Donner store location, the company stocks about 240,000 titles on 55,000 sq/ft and receives about 8million units each year. In a ‘traditional’ store, stock takes require the closing of the store for 2days. The new store remains open and the inventory is counted in hours. Theft and shrinkage have been reduced because the company accurately tracks items from receipt through purchase and the RFID tags acts as a theft alarm. (Privacy has been addressed by deactivating the tag on purchase). In a presentation at Frankfurt last year, the company estimated that they save over $250,000/yr just from stock take efficiency alone. Evidence also suggests that average revenue per customer has increased 6-8%, due to better inventory information at store level.
During implementation, the company needed to address minor issues such as metal shelving that interfered with RFID reception, metal or part metal packaging on books, CDs and DVDs; and, in some cases, location of a title was not specific enough – covering two bookshelf bays for example. These issues appear minor and, post-implementation, the store experience for shoppers is considerably enhanced. Employees and customers can identify with certainty an item and its location in the store. More integration with their web site and with store promotions and bundling is planned. The company hopes to have RFID implemented store-wide by the end of 2008, with added improvements (such as RFID enabled shelving) which could eliminate stock taking entirely.
The prospects for success in the US appear muted for two reasons. Firstly, well-intentioned industry volunteers seem to be fixated on defining the data that may or may not ride along on the RFID chip. In the BGN implementation, the chip only contains a unique number. At the point when the RFID chip is applied to the item (by BGN), the chip number and associated meta-data are married up in the BGN product database: Taken at face value, this approach appears more flexible, cheaper and faster. (For example, what happens if data ‘formatted’ on an RFID chip is inaccurate? The same mistake could be rectified once in a database rather than having to recall all RFID chips and rewriting the information). In the US, the initiative may move faster if we just define the syntax for the ‘dumb’ RFID number. Secondly, no one in the industry appears to want to take the first step but, as the BGN example shows, significant advantages could accrue to the company that goes first. In my opinion, this is most likely to be a retailer and, perhaps, should be a retailer, because the business case seems to be more obvious.
In looking to the future, BGN hopes that tags will be applied at the bindery and supply-chain partners can adopt the technology when they are able. In the US, we don’t seem to have reached that point of mutual desire. In an industry where a half- point gain in operating margin is hard to find, one would think that capital investment supporting RFID implementation at store level would be a no-brainer- especially given the example of BGN.