Big data trend now being applied to managing human resources

A growing number of human resources executives are starting to dig deep into computerized statistical data on employees, to make decisions regarding salaries, promotions, and even benefit programs.

It's a trend that excites some and worries others.

Andrew Martin, who oversees human resources at the Joey chain of restaurants, which operates 30 locations in British Columbia, Alberta, Manitoba, Ontario and Washington State, believes the use of so-called "people analytics" is great for companies and for employees.

"Have you seen the movie Moneyball?" Martin asks, referring to the 2011 film that starred Brad Pitt as the general manager of the Oakland Athletics baseball team, who pioneered a statistical approach to choosing players.


Instead of baseball's RBI's or earned run averages, HR departments track data such as absenteeism, salaries, sales figures, and are increasingly looking at ways to attach numbers to qualitative measures that come from performance reviews and assessments from colleagues.

"It's brilliant," Martin says of his favourite movie. "And you know, people say it's complicated, you can't reduce people decisions to black and white. It's not really true. Performance is performance."

Some won't like it

Business ethics professor Chris MacDonald is on the other side of the issue, questioning the use of statistics for decisions about people's place and success in the workplace.

"Statistics have their uses, but you don't want a number to sum up the whole employee- employer relationship," says MacDonald, who works at Ryerson University's Ted Rogers School of Management.

"It may well be a self-defeating mechanism if you're trying to manage people by being incredibly precise by measuring their performance. If that measurement process becomes demoralizing for staff, you've shot yourself in the foot."

So called "big data" techniques have been embraced quickly by sales, finance and marketing departments eager to exploit trends that will help generate more revenue for their organizations. The move towards harnessing computer power has been slower in areas that don't generate profit, such as the personnel department. But that's beginning to change.

Numbers tell the story

"I was in a meeting several weeks ago in San Francisco and we had eight PhD statisticians, engineers, and computer scientists together, all working on people analytics for their companies," Josh Bersin wrote in Forbes Magazine recently. Bersin is one of the best-known American consultants on this trend.​

In an interview with CBC News from his office in Oakland, Bersin says the use of hard data can actually make the workplace more fair.

"We all have biases based on our experience. You're basically relying on a data set that you personally experienced that may be not 100 per cent relevant. When you look at data, you take emotion out of it."

Martin says the Joey restaurant chain has used data to design a more fair way to pay its chefs and general managers. Previously, salaries were based on net sales, which are often related more to location than to employee performance.

"People naturally want to fluff up the people they like, and they start talking about how great they are, or how pretty their girlfriend is. When you grab the actual numbers, it's stopped people in their tracks. They weren't expecting that," he says.

The chain measures a variety of indicators for its chefs and general managers. Restaurant cleanliness, wait times, and revenue growth, along with assessments given by other employees, are used to determine a ranking.

Martin says he intends to expand the system to other job categories, boiling a collection of scores down to just one number.

"For me, you can look at a whole whack of numbers but it gets lost until you can bring it down to one number," he explains.

I'm more than a number

Many Canadians could find it offensive to have themselves and their value to their employer reduced to a single number. And there's also the matter of what's acceptable in our culture, says Dr. SeonaidCharlesworth of Knightsbridge Human Capital Solutions. She works primarily with senior executives, using an algorithm to assign numbers to a range of leadership abilities.

"I've done a lot of work in the U.S., and when I returned to Canada I was shocked at the number of leaders who felt it would be counter-cultural to look at whether people had different abilities, and to make decisions about who to hire, promote or invest in, based on their underlying ability. They had a sense people were different but didn't want to go there for fear of disengaging the broader group," Charlesworth says from her office in Edmonton.

"In the U.S., they have the mistaken notion everyone is equal. In Canada, we have the mistaken notion we can force everyone to be equal," she added.

Analytics present a raft of limitations, says Jerry Gaertner, a professor of computer science who is also the co-owner of an analytics company. "What's the unit of measure for initiative? Or enthusiasm? It doesn't exist," he points out. He also says it's next to impossible to remove bias, even by using numbers.

"The data remains has to be interpreted by a person and any time you get a person interpreting data you get bias. The data shows you a pattern. You still need to identify what the pattern says and how to use it."

Air Miles joins the trend

But that hasn’t discouraged LoyaltyOne, the company that runs the popular Air Miles program. The firm uses analytics to attract and keep "top talent," according to vice-president of talent management Diane Dowsett. She says that by pulling data on the source of referrals, and linking that to how the new employees performed in the job, she was able to see that the best hires had been referred by people already working at LoyaltyOne. The next step was simple.

"We increased our employee referral incentive program," she explains. "Because we run Air Miles, we can offer a substantial amount of Air Miles as a reward." The dollar value of these rewards ranges from $2,000 to $10,000, depending on the level of specialty involved in the job.

Bersin compares worry about people analytics to the fear that developed almost 10 years ago when consumers first learned of "cookies," the packets of data that track your destinations on the internet.

"We were all worried because we thought Google or Facebook would hurt us," he says. "But really all they did was make their sites more useful. Now Amazon recommends things I actually want. I think that's what is going to happen in HR. Analytics will really help business managers do a much better job of hiring the right people and managing their workforce in a better way so they have a better place to work."