RFID, Fast Follower Strategies Don't Mix

Concurrently, many layers in the enterprise technology stack will require RFID impact assessments. These assessments include activities such as environmental surveys at each location to weed out sources of interference and to determine how RFID will perform, from that specific location all the way to the back office and business intelligence applications that run the enterprise. Furthermore, some of these assessments are required even if the retailer or supplier conducted their pilots in an RFID testing facility. Here is where your system integrator -- Accenture, BearingPoint, Capgemini, IBM Business Consulting Services, Wipro -- can earn their salt.

However, experience gained from a testing center can enable enterprises to shortcut some issues. These testing centers can contribute from experience and shorten some investigations such as the effects of liquids and metals on tag readability, as well as tag placement on cases and pallets.

Additionally, we believe it will take enterprises two to three years to take on all the issues within the technology stack.

Issues they must deal with include installation of readers, gates, light sticks, device management, case/pallet configuration, application server and networking upgrades, and integration with enterprise applications. Therefore, IT organizations should garner resources from several areas to support RFID pilots and assess the impact to the IT portfolio holistically. RFID pilots will touch in-store systems, warehouse management, and back office systems and require physical infrastructure requirements such as power and network connectivity.

Furthermore, application infrastructure components such as application servers will require upgrades. Vendors including IBM and Oracle have launched new products to support these requirements, Meantime, SAP has incorporated these functions into the AutoID infrastructure part of the NetWeaver stack. These upgrades will enable RFID data aggregation and filtering of events from non-events, as well as control of RFID components such as readers, both handheld and gates, light-sticks, and alarms. Therefore, IT organizations should assess and address the impact to applications early instead of focusing on the limited scope of the tag testing pilot and determine how to leverage RFID data for business intelligence and data warehousing applications.

IT organizations should use 2005/06 as an RFID learning period (if they're not already under way with the technology) and prepare the enterprise for RFID adoption. This will enable the enterprise to put in place the adaptive infrastructure needed to support RFID's sensing capabilities and integrate with systems necessary to create the proper enterprise response.

Another question companies are asking is this: Are the RFID stress factors bigger for suppliers than retailers?

RFID initiatives do not disregard retailers since stores and distribution centers must be outfitted with readers and antennas. Retailers will need to carry the costs for this outfitting. Furthermore, many of the suppliers' stress factors listed above are the same as those of retailers. However, suppliers are bearing the burden of investing in RFID earlier than retailers because of pull created by major retailer mandates. Indeed some of the lessons learned by the suppliers will be passed on to the retailers (such as tag placement issues for cases and pallets) with no costs to the retailer, and many suppliers hope that this will be offset by incentives such as preferred supplier status and information on how their products are responding to consumer requirements.

Retailers and suppliers that wish to be fast followers should begin RFID piloting and assessment of the impact to the total IT portfolio. If not, a fast follower strategy will not be possible due to the depth at which RFID initiatives can reach.

These enterprises should begin with a barcode replacement methodology and then progress into the utilization of RFID for internal and external process and data sharing improvements. This incremental approach will enable enterprises to spread out the costs of RFID projects over several fiscal years instead of attempting to react within a single fiscal year and keep pace with the competitive market.

Bottom Line: The evolution of RFID from experimentation technology has begun and retailers and their suppliers should not let this

technology get too far of ahead of them due to large infrastructure and evaluation costs.

Gene Alvarez is VP of technology research services at the Meta Group. He has 20 years of IT experience in business-impact assessment, vendor management, project management, software development, and delivery of complex business applications. He also has held positions with Nine West Group, KPMG Peat Marwick, New York Power Authority, and AT&T Communications..