Datacenters still a boys' club, staffing shortages may change that

Fifth of server warehouse operators polled didn't employ a single female worker

Datacenter operators' investments in inclusion and diversity have done little to shift the balance of workers in the historically male-dominated field, an Uptime Institute report found.

Three quarters of datacenters surveyed by the analyst group said women make up roughly 10 percent of their workforce, while a fifth of respondents said they didn't have any female workers at all. And while Uptime notes that the ratio has improved from five years ago, gender disparity hasn't shifted by much.

According to analysts, hiring more women could help to address staffing shortages, which it projects are only going to get worse over the next few years.

"A growing number of unfilled positions coupled with a low and stagnating proportion of women workers suggests the datacenter industry still has much work to leverage the untapped potential of the female workforce," the report reads.

Finding qualified staff to keep datacenters running has been an ongoing trend, but it has gotten worse over the last few years as companies invest in more computing power. Today, 54 percent of datacenter operators cite staffing as their primary challenge. That's up from 47 percent in 2021, and 38 percent in 2018.

And when datacenters can find staff, they're having an equally hard time keeping them. According to Uptime, 42 percent of datacenter operators said they were having trouble retaining staff, in part because they're being poached by competitors. That's up from 17 percent of datacenters five years ago.

Making matters worse, Uptime says that the staffing shortage is likely to get worse over the next few years. Analysts report that a large portion of the datacenter workforce in North America and Western Europe are inching closer to retirement.

"This means a significant portion of the existing workforce will retire concurrently — leaving datacenters with a shortfall on both headcount and experience," they wrote.

As a result, Uptime anticipates that staffing will remain a major challenge for datacenter operators for the foreseeable future but notes that employers may find success attracting and retaining workers through training and mentoring programs.

Efforts to increase automation, particularly those systems employing machine learning or AI to reduce the burden and allow a smaller number of staff to work more effectively, are often touted as the answer to these challenges.

Networking vendors in particular, continue to push the idea that AI will eventually automate network management and remediation. However, datacenter operators appear to be lukewarm on the idea.

On one hand, Uptime says more datacenter operators are expressing confidence in AI, with 57 percent saying they'd trust an "adequately trained machine learning model to make operational decisions." The increased confidence may be the result of high-profile advances around generative AI, like ChatGPT, Bing Chat, and other large language models.

However, Uptime reports that most datacenter operators don't expect such a model to be ready within the next five years and only about half of respondents expect AI to allow them to reduce headcount.

So, on the bright side, if an AI model ends up rendering your job irrelevant, there's a good chance you can find work keeping the datacenter where it's housed running. ®