While big data certifications in skills like MongoDB or Cloudera were in top demand over the last two years, pay premiums began to erode mid-2014 as companies started taking stock of their projects and determined they weren't returning quite as much value as expected, according to Foote.
Non-certified big data skills (in categories like data mining and master data management, for example) took a bigger hit the last year, showing measurable declines in pay over the past three quarters.
Foote and other analysts remain bullish on investments in big data skills into this year and later, but most believe companies are now more realistic about the staffing and technologies needed in order to ensure success.
"Companies are realizing that they just can't spend a lot of money and have big data work," Foote explains. "There needs to be a culture of transparency in the organization and a lot of sharing going on."
In practice, big data efforts at many traditional companies have been hamstrung by hierarchical structures, inadequate data governance and management's distrust of data-driven decision making, Foote says. Before big data specialists have a chance at success, companies need to hammer out these management issues, he adds.
There's also the issue of maturity when it comes to analytics. "From an organizational and cultural point of view, most companies aren't doing a good job with the analytics they already have, so how do you expect to do big data," he says. "You can't get there simply by hiring the best and the brightest." As a result, the premium organizations were previously putting on those certifications has softened a little, he says.