"A decision was made to continue to supplement with independent contractors to address these smaller, one-off projects and to fill service gaps," says Infante, who declined to identify the chemical company. "The arrangement quickly became difficult to manage and the outsourcer was removed."
One of the main causes for the failed outsourcing engagement, which resulted in both lost time and increased costs, were underestimating the effort needed to manage the outsourcing relationship, Infante says.
"Where the client thought it would take one or two individuals to manage the interface with the vendor, it actually took many more of the company's resources," he says.
In addition, the client's interpretation of what it was buying -- based on what it was told by the outsourcing vendor's sales team -- wasn't interpreted the same way by the vendor's implementation and service teams.
How could these kinds of problems be avoided? "Know thyself," Infante says.
"The hardest thing for many businesses to do is properly assess their real needs and, more importantly, their current position regarding the state of IT. Properly defining these two components will ultimately determine what can/should be outsourced and which company best matches their needs."
Todd Taylor tells tale of an outsourcing arrangement gone awry due to differing needs among business divisions.
Taylor, an attorney at Moore & Van Allen, who focuses on outsourcing and other technology issues and is familiar with the case, says the arrangement involved a large, multinational corporation and a large, highly regarded technology services company.
The multinational had multiple business units that provide services for consumers and businesses, Taylor says. It outsourced portions of its network operations and infrastructure to the technology services company to support multiple divisions, each of which had different needs.
The client spent time with the service provider coming up with service-level arrangements. But the service-level metrics were typically configured on a clientwide basis.
"In other words, in determining whether a service-level metric was met -- or not met -- performance was judged for the client on an enterprisewide basis rather than on a division or line-of-business basis or a specific service element basis," Taylor says.
In many cases, certain business units or functions weren't receiving acceptable services, and this potentially hindered the ability of those units to meet their customers' needs on a timely basis.
"The client had little contractual ability to demand correction of the problems, as the service provider was generally meeting the service-level metrics when such metrics were measured on an enterprise-wide basis," Taylor says. "Even when service levels were not being contractually met, the penalties often were not meaningful enough to incentivize the service provider to change behavior."
Sign up for CIO Asia eNewsletters.