Dorfman Pacific execs also wanted to make certain that the wireless component, a critical piece of the overhaul, would work in their warehouse. They hired Texas Bar Code Systems to conduct a radio frequency study to see if the wireless signals would play nice inside the facility's concrete walls, steel doors and metal racks, and to identify the best wireless access points.

The IT infrastructure gluing this all together also needed a makeover. Dulle had to revamp the ERP system he inherited as well as select, install and integrate a new warehouse management system that had wireless capabilities and could sift through the company's warehouse data and shipping information. He also ripped out the old networking, cables and switches, and upgraded to the latest and fastest network gear and fibre.

After hiring Symbol Technologies for the wireless networking equipment, HighJump Software for the warehouse management system and Zebra Technologies for the bar-coding equipment, and selecting an integrator (RedLine Solutions), Dorfman Pacific could see its future warehouse. The backbone would be a wireless local area network (with 802.11 connectivity) that utilized 15 wireless access points (APs) spread out over the facility that could simultaneously run the 802.11a/b/g bands on each AP. Warehouse staff would have 40 mobile and fixed-mounted computers on the forklifts at their disposal. The devices were kept simple: "[It was] all F1, F2, press this key, press this button — no mouse and no pen," Dulle says.

In the back office, IT was set to roll out an upgraded ERP suite and the new warehouse management system that would direct the picking, packing and shipping processes using specialized logistics software. Data from the software would appear on the forklift's mobile device and tell a picker where to go, what to grab and where to bring it in the most efficient way. Paper would be a thing of the past.

Learn from Mistakes

In late 2004, executives and the project team had to sell the change and set the expectations for the March 2005 go-live date for the folks on the floor. Executives held group meetings and one-on-ones, formed focus groups and selected project leads to get the message out. "We said: We're going to make your lives different," Dulle recalls. "We're going to make it harder to make mistakes. We're going to make it more efficient. We're going to help you to do a better job." Everyone got on board.

In the race to meet the March deadline, as the pressure to deliver the new system pushed everyone to work nights and weekends, and as the opportunity to deliver the first reduction in temporary labour costs loomed, Dorfman Pacific hit the brakes. Despite everyone's efforts, the new system and processes just weren't ready. "We pushed this very, very fast. We were trying to drive operational costs down, and we let that colour our judgement," says Dulle.

The go-live date moved to July. It was a costly decision. Dorfman employed the largest amount of temporary labour in March and April, and executives were counting on this first reduction in costs. However, the extra time allowed employees to work through lingering issues with the IT systems and the new shipping and inventory flows. For example, the warehouse management system required a different floor configuration and new ways to pick, pack and ship products. It also placed more emphasis on the process of expediting the merchandise no matter how many other products were a part of the order and were, most likely, going to the same customer.

For bigger retailers, receiving 20 boxes instead of five wasn't a problem. But for the smaller stores, receiving 20 boxes instead of the usual five made a huge difference because of the accumulated freight charges. The delay let Dulle and the team reconfigure the software to fix the problem and keep smaller customers happy.

CEO Highsmith says the decision to delay was based on how the premature rollout might have a negative effect on Dorfman Pacific's customer base so close to peak season. "The technology embrace has to have a positive impact on customers. If it makes it more efficient here but has a negative impact on the customer, whatever gain you make you lose on the other end."