The Chile-based retailer Falabella completed a successful apparel RFID pilot project last year, and is cautiously forging its RFID strategy for the future.

 

A 118-year-old retail giant with annual sales of $4.3 billion and 70,000 employees, the Falabella group operates department stores, home improvement stores, supermarkets, a travel agency, insurance agency and bank, including its own credit card business. The privately held firm's operations are spread throughout Chile, Colombia, Argentina and Peru.

 

Falabella staged the RFID pilot at one of its department stores in Santiago, where it attached RFID tags to select men's clothing from the Lacoste brand and a popular local fashion brand. The tagged apparel included men's tailored garments as well as suits, pants, shirts, T-shirts, sweaters and belts - all goods that Falabella perceived to be of high value to the store's customers.

 

The retailer did not involve apparel vendors in the pilot.

 

A well-planned pilot
Falabella waded into RFID cautiously. The firm kept tabs on Generation 1 RFID reliability reports and "was not impressed," says Cristian Astaburuaga, RFID project team leader at Falabella.

 

Systems At A Glance

RFID Business Process Consulting & Integration: IBM

RFID Tags & Mobile Carts: Paxar (Avery Dennison)

Hand-held & Mobile RFID Readers: Motorola

RFID Software: OATSystems

 

But later, after reviewing evidence of more reliable Gen 2 data and attending RFID demonstrations in the United States, Falabella decided it was time to act.

 

In early 2006, the retailer outlined three goals for an RFID pilot project: 1) to conduct a daily inventory count of RFID-tagged apparel; 2) to achieve at least 98 percent reliability of RFID tag readability and data accuracy; and 3) to build a business case for RFID.

 

The retailer committed to weekly RFID pilot project meetings between employees representing its technology, business and store operations. "The synergy between the different departments was very important," Astaburuaga says.

 

Even as it moved forward with its RFID trial, Falabella was worried about the reliability promised by Gen 2 RFID. Before investing in its live pilot, the retailer could not easily obtain an RFID reader to test the technology. "We really had to conceive it on paper," Astaburuaga says.

 

Selecting RFID technology suppliers also was challenging. The retailer spent three to four months interviewing 10 vendors, and found that half of them had little to no actual RFID experience, says Andrés Arancibia, Falabella's research and development chief.

 

The company ultimately chose IBM for RFID business process consulting and as the lead technology integrator for the project. Falabella selected Motorola's MC9090 hand-held RFID readers for the receiving process and built two battery-powered mobile carts using both a fixed reader and antenna readers for inventory counts. The retailer purchased its RFID tags and other mobile carts for the pilot from Paxar (Avery Dennison), and chose OATSystems for RFID software.

 

Early results: pleasant surprises
The four-month pilot ran from January through April of 2007. The first couple of months involved a lot of integration troubleshooting to access the RFID data properly and debug the systems.

 

But then the benefits began to show - quickly and abundantly.

 

One of the three goals for the pilot was achieved right out of the gate: Readability of the RFID data was 99.7 percent accurate.

 

With confidence in the data, Falabella moved to tackle another goal. It began conducting daily inventory counts of its RFIDtagged apparel, attaining 98.4 percent inventory accuracy.

 

In each daily cycle count, there were about 2,500 RFID-tagged garments, depending on stock levels in the store. During the course of the pilot, Falabella tagged a total of 7,000 apparel items with RFID tags.

 

Every morning before the store opened, a Falabella associate rolled a mobile RFID reader cart through the apparel department, automatically collecting RFID signals emitted from the Paxar tags. This data was fed into the OATSystems software for report generation.

 

The RFID-enabled inventory counting was one of the most compelling achievements of the pilot because of its clear timesaving benefit. Falabella proved that an inventory count could be done with RFID in a matter of hours. Done the traditional way, that count would require 50 employees per store to work for two full nights while the store was closed - a major undertaking only possible to schedule monthly, if not quarterly.

 

An opportunity for "new conversations"
The pilot focused on gathering data about three key activities in the Falabella store: receiving, restocking and checkout.

 

Knowing it had reliable data, the RFID project team began generating regular reports for business and store managers.

 

In fact, there was so much data flowing out of the pilot that Falabella struggled at times with how to use it, Arancibia says.

 

For instance, while both the RFID project team and store managers knew that the RFID data could give Falabella much greater insight into shrinkage, they initially had different views of the best way to define shrinkage and analyze the data.

 

The sheer speed of data delivered by RFID broke many paradigms for Falabella. For years, the firm's measure of shrinkage had been closely tied to the time-consuming traditional inventory count procedure. With newly accessible data, Falabella had to think about its inventory in brand new ways, says Arancibia.

 

"We started some conversations we had never had before€š" to arrive at new KPIs not only for shrinkage but also for other important issues, such as determining how long garments stayed in the backroom, on the sales floor or in-transit between the two, he says.

 

Tough decisions about how to proceed
With a successful pilot under its belt, Falabella is working to forge a pragmatic plan for a larger-scale RFID implementation. While he could not share the amount Falabella spent on its pilot, Arancibia says it was expensive.

 

The company must be confident in the financial assumptions behind a larger RFID project before committing to fund it, he says.

 

The RFID project team developed an RFID business case with ROI estimated within 3.7 years, but is re-evaluating the proposal with the leadership team.

 

"I'm pretty sure we'll proceed, but it's a matter of how to best proceed," Arancibia notes.

 

In the meantime, the project team is developing tools for smaller-scale tests so that it will be ready when the funding comes, he says.

 

"RFID isn't as hard as it seems after the first pilot," Arancibia concludes. "With the pilot, we developed a lot of confidence in the technology, what it can do and what we can do with it." Apparel