The Economics of Data Center Staffing

How many employees does it take to run a data center? The Green Data Center blog explored the issue last week when it looked at hiring for new data centers in central Washington and found that Yahoo (YHOO), Ask.com, Intuit (INTU) and Microsoft (MSFT) have hired a total of 180 workers for their facilities. Microsoft, Yahoo and Intuit will each have 50 workers at their data centers in Grant County, while Ask.com is adding 30.

Blogger Dave Ohara contrasted those numbers with Google's standard figure of 200 new jobs for each data center project, and wondered about the discrepancy. "Something seems strange that four big data centers in the Columbia Basin have less employees than one Google data center," he writes. "What are those 200 people doing? Are they maintaining and building Google's custom servers?"

Data center employment often comes up in discussions of economic development. Many communities are eager to attract data center projects, but struggle to define the economic benefits of these facilities. Jobs have always been the primary benchmark by which economic development projects are measured. Incentive packages offered by state and local governments are often based on the number of full-time jobs created by a new business.

This model doesn't work out well for data centers, which are typically highly automated, allowing a small number of workers to operate a large facility. A new data center can bring a large capital investment into a town, but create a much smaller number of jobs than a factory or office property of a similar size.

Some local officials like this scenario, since data center projects can attract a large ratable without adding significant traffic or large numbers of children to the local school system - two areas where large new businesses can create costs and headaches for municipalities. Others are less comfortable offering incentives to companies for buildings that will house electronics rather than people. This debate recently played out in a city council debate about a data center campus in Old Bridge, New Jersey.

As we've noted many times, comparing data centers is a tricky business, as facilities differ in ways that make apples-to-apples analyses difficult. That's definitely true with job creation, as data center projects generate large numbers of temporary jobs during the construction phase. As an example, Microsoft's data center in Quincy, Washington had as many as 500 workers on-site at a time during the construction process, but now employs 50 full-time staff to man the center.

As for Google (GOOG), they don't talk about their data center operations (see the Fight Club Rule of Data Center Secrecy), so it's hard to know precisely what those 200 jobs represent, and why they employ four times as many workers per facility as Yahoo or Microsoft. Perhaps we need a new data center efficiency metric: employees per square foot.