Missing the potential of inventory optimization

Inventory optimization remains a popular type of application in distribution these days, but most companies do a poor job of making that software work for them. That's the opinion of a leading expert in the field, Shaun Snapp. Now an independent consultant, Snapp worked as an analyst on an early inventory optimization software implementation in the late '90s. He has written two books on the subject: Inventory Optimization and Multi-Echelon Planning Software and Supply Chain Forecasting Software.

Although inventory optimization (IO) software is a relatively recent innovation, the science behind it is not. Snapp says the Air Force developed the mathematics behind inventory optimization software in 1958. Back then, however, computers were not capable of running the algorithms. "It took decades before computer power was able to keep up with the software," he says.

According to Snapp, the use of inventory optimization programs started to take off in 2006. Since that time, a number of companies have implemented IO applications, which calibrate stocking levels across the supply chain. "I'm amazed at what inventory optimization can do," says Snapp. "It can set service levels at detailed levels, and let the optimizer [in the software] decide where to put the stock. There's no debate about the theoretical value of inventory optimization."

Many companies have taken notice of the value of inventory optimization software. Nearly a quarter of the respondents to a November 2012 DC Velocity survey said they planned to buy inventory optimization software in the coming year, making IO software second only to warehouse management systems on their "to buy" list. (See "Getting the big-picture view of inventory," December 2012.) Some of the top vendors of this software, according to Snapp, are Logility, JDA, Manhattan Associates, Servigistics, SmartOps, and ToolsGroup.

Although companies appear to be sold on the software's value, many seem to have trouble getting the most from the technology. Snapp says there are a number of reasons for that.

For starters, there's the comfort issue. In many of the cases where inventory optimization has not met corporate expectations, Snapp suspects that the software vendors did not adequately explain how the application works. That makes it tough to get buy-in at the user company. Because they don't comprehend the logic behind inventory optimization, company executives don't trust the software's recommendations to reduce stock levels at specified warehouses and plants. "Somebody will look at the output [from the software] and say, 'I'm not going to do this. We're used to maintaining a certain amount of inventory,'" Snapp says. "If you don't trust it, you will keep the old systems."

What users have to understand, Snapp says, is that the software takes a broad, interconnected view of inventory holdings in the supply chain. "Stocking locations used to be viewed as islands," he explains. "But when you use this software, one location is making its decision on stocking based on what another location is doing."

In other cases where the software fails to live up to its billing, the firm implementing the software lacks the right personnel for the job, Snapp says. A successful implementer needs two skills, he explains. The first is the ability to understand and troubleshoot the application. The second is the ability to "socialize the solution" within the company. "It is normally not feasible to expect both skills from the same person—or even that one person be quite good at both," he says.

Despite those issues, Snapp still believes that more companies should take advantage of this software tool. IO holds great potential to boost efficiency and reduce costs, he says, but only if users are willing to put some effort into the deployment. "Implementing inventory optimization is more complicated than most people think," he says. "It needs to be put in better, and people need to become better educated on how to use it."