Human Resources Tentatively Tries Predictive Analytics
Knowing the probability of important employee events before they happen can have a big bottom-line impact.
5 Big Wishes For Big Data Deployments
(click image for larger view and for slideshow)
What's the probability that employee X will leave in two years? Could predictive analytics supply an answer?
Accurately forecasting what any individual employee will do in the future is at the bleeding edge of the market, say human resources experts. What's more common, and on the rise, is using analytics to better understand the patterns of large collections of employees, such as in a call center.
"Statistical techniques used for prediction tend to work with larger numbers," David Gartside told InformationWeek in a phone interview. Gartside is managing director responsible for HR offerings and capabilities within the Accenture Talent & Organization practice.
In call center operations involving thousands of people, such analyses are being used today, providing, for example, predictions about the percentage of workers likely to leave in a month.
"If you have a good view of this, you can plan accordingly, ramping up or down recruiting," Gartside says.
Three things are driving the use of predictive analytics in HR, Gartside told us. First, HR departments are getting much better at using operational processes and technology with an eye toward collecting good-quality data to make better decision-making.
"The second piece is social data," he said, referring to the inclusion of both external and internal data. These rich data sources didn't exist even a few years ago.
Finally, he notes, vendors of HR solutions are increasingly building analytics into their core platforms.
But predicting an individual's future actions -- think Minority Report-style "precrime" -- raises a number of largely unanswered legal and ethical questions, too, which explains why HR organizations have been pursuing this application of predictive analytics with a great deal of caution.
In the context of NSA spying revelations and other privacy concerns, "people have a heightened sensitivity" about surveillance, said Mark Berry, vice president of Human Capital Analytics and Reporting at ConAgra Foods, during his presentation at the Predictive Analytics Innovation Summit in Chicago earlier this month.
Nevertheless, ConAgra Foods, which has only just embarked on some HR analytic programs, hopes the work will help it plan better and improve business outcomes.
"We want to know our employees as well as we know our customers," Berry said, adding that the company has already developed a number of safeguards for what types of employee data it will and will not collect, as well as assess the impact, both positive and negative, of the project before proceeding.
Where to start But how accurate is predictive analytics when it comes to forecasting individual employee events, such as a key vice president of sales quitting without notice?
"It is a very difficult science," Accenture's Gartside says. And like ConAgra's Berry, Gartside urges companies to think about how these systems will be regarded by employees and the marketplace. Make sure these programs aren't just cognizant of what's legally allowable, he cautions, and make sure they are aligned with the culture and company brand as well.
Asked for advice on how to get started with predictive analytics in the HR function, Gartside offered the following:
Start with a business problem, such as service quality, being impacted by employee attrition.
Do a pilot with existing data and capabilities. "See if these analytics have value, and don't wait for the data to be perfect."
Finally, put in a technical infrastructure that can make this kind of analysis repeatable and easy to do.
Will advanced, data-driven approaches to employee performance and outcomes become standard? Gartside thinks so.
"Look at how many people have a job title with 'talent analytics' in it. This title didn't exist two years ago," he said, adding that the Fortune 250 are carving out this executive role, "because people are finding value in it."
Emerging software tools now make analytics feasible -- and cost-effective -- for most companies. Also in the Brave The Big Data Wave issue of InformationWeek: Have doubts about NoSQL consistency? Meet Kyle Kingsbury's Call Me Maybe project. (Free registration required.)
InformationWeek Tech Digest, Nov. 10, 2014Just 30% of respondents to our new survey say their companies are very or extremely effective at identifying critical data and analyzing it to make decisions, down from 42% in 2013. What gives?