The benefits of insourcing data center operations

Though outsourcing data center operations works in some scenarios, insourcing might still be the best solution for many IT shops.

Outsourcing may be in, and insourcing may be out, but before pushing data center operations entirely to the cloud, consider the parallels in other industries.

Boeing's newest passenger plane, the 787 Dreamliner, has been in the news for equipment failures. Batteries catching fire, electrical problems and fuel leaks are among the problems that have grounded the planes, on top of the serious delivery delays that customers -- the airlines -- have experienced.

Many analysts have blamed the extensive outsourcing. As much as 70% of the design and assembly was outsourced, unintentionally driving knowledge and profits to the suppliers and away from Boeing. The result is a product that is billions of dollars over budget, late, malfunctioning and grounded by the Federal Aviation Administration.

Other companies are doing exactly the opposite with operations, bringing design and manufacturing back in-house and back into the United States from China. GE is building appliances again in its long-shuttered facilities in Kentucky. Its water heaters cost less, are better quality, and don't incur long-distance freight delays and costs. A big reason to move is designers can talk directly to the implementers such as the welders who assemble the heaters' internal plumbing. They have pinpointed problems and made design adjustments that result in higher quality and faster assembly.

The upsides of insourcing

There are many parallels between what the IT industry is undergoing and the outsourcing transformations in manufacturing industries during the 1980s. All we talk about now is moving to the cloud, pushing workloads into far away data centers, copying files to DropBox and loading customer data into, after which we wheel our servers out the back door. This is today's outsourcing. Companies reduce their data center footprints and staff, and as staff members leave, they take hundreds of person-years of knowledge with them.

GE discovered that having a feedback loop between the designers and implementers meant better, less expensive products. What about the feedback loop between developers and system administrators, or between architects and users? Can that loop even exist when most of the system design and implementation is done by a supplier, such as or Dropbox, with no chance for input? Will the new IT suppliers do a better job than in-house techs?

I don't mean to be alarmist. The cloud offers new choices and potential freedom from the brutal enterprise software we've been stuck with for years. Absolutely everything in life is a tradeoff. But it's important to note that as CIOs adopt a herd mentality about cloud computing, we need to consider what we might lose. Are we pushing critical IT and business knowledge out the door -- possibly to our competition -- as we downsize our server room? Are we gaining agility in the cloud or losing it with the rigid one-solution-fits-all approaches of cloud providers? Is there an exit strategy from the cloud? In the long run, would it solve more problems and be cheaper to hire knowledgeable staff, and keep IT operations in-house? As everybody else outsources, could insourcing IT operations be a competitive advantage? These are tough questions. I just hope we have answers before the cloud turns us all into 787s.

Let us know what you think. Write to us at [email protected]

Dig Deeper on Colocation, hosting and outsourcing management