McDonald’s Canada held its national hiring day on April 11, and received more than 37,000 applications. The burger chain, which employs about 80,000 people across the country, hired 5,000 new workers, many of whom were needed to replace others who had quit or otherwise left the company.
Hiring and training new employees is expensive, and most corporations try to keep staff turnover, or “churn,” to a minimum. But it’s a particular challenge in the fast-food industry due to its heavy reliance on young part-time workers. While McDonald’s has largely accepted this as a fact of life—its hiring slogan is “for a year or a career”—sorting through all those applications still represents a supersize challenge for store managers trying to determine which workers, many with no previous experience, will be most likely to succeed. And so, in 2006, the burger giant turned to consulting firm Aon Hewitt in the hopes of making the whole process more scientific. They came up with a 15-minute survey, constructed by McDonald’s and Aon’s team of industrial psychologists, that’s administered to applicants and later analyzed by computers. The software suggests which applicants should be contacted for an interview and helps inform managers’ questions. The approach sounds cold and impersonal, but McDonald’s says it works. “It took a lot of the guesswork out and makes it a lot easier to identify those individuals that have the aptitude and willingness to learn,” says Len Jillard, the “chief people officer” for McDonald’s Canada. “Turnover has been reduced.”
McDonald’s is not the only company now outsourcing traditional HR functions to computers. And the machines are doing more than just helping managers decide who to hire. In a blog post, Josh Bersin, the president of consulting firm Bersin & Associates, described the trend as “Moneyball comes to human resources,” referring to the popular book and film about Oakland A’s general manager Billy Beane and his use of sophisticated statistical analysis to build a winning baseball club. Bersin cited firms as diverse as home improvement retailer Lowe’s, Credit Suisse and Accenture as among those that now employ analysts who spend all their time “analyzing data about their own people, to identify what precisely makes up a high-performer, and how to best replicate this performance in the workforce.”
It’s all part of a larger shift toward the use of data analytics, or “big data,” in the corporate world. While crunching massive databases in search of insights into consumer behaviour is fast becoming standard operating procedure (Amazon’s product recommendation engine is a good example), the notion of using the same tools to manage a company’s own employees is more controversial, mainly because it’s not an exact science—which, as baseball fans all know, was something Beane discovered when his stats-stacked team failed to win a championship.
The push to modernize HR departments in recent years has been accelerated by the turbulent global economy. In an era of high unemployment rates, many big companies now face a McDonald’s-like river of applications every time they post a job opening. “There are just so many people who are qualified, along with those that aren’t a good match but who are still applying for these positions,” says Paul Barsch, the director of marketing at Teradata, a data analytics company based in Ohio. “Companies need a system on the front end of this hiring process.”
Sensing an opportunity, firms that sell business-oriented software are bulking up on HR-oriented tools. Germany’s SAP bought SuccessFactors, an employee management company for US$3.4 billion in December. In February, Oracle bought Taleo, a human resources software company, for US$1.9 billion. And IBM paid US$1.3 billion for Kenexa, another talent management and HR software-maker, in August.
The potential market for such products is huge. A recent IBM survey of 1,700 chief executives found that 71 per cent said human capital was the most important factor in maintaining a competitive advantage. Still, despite the increasing popularity of big data in other areas of business, another IBM survey of CEOs found that only slightly more than one-third of respondents felt their companies were effectively using analytics to make strategic decisions about their workforces. Neil Crawford, a partner at Aon Hewitt, says the apparent disconnect results from the relatively low position HR departments have traditionally occupied on the corporate totem pole. “There’s not a lot of capital being spent on it,” he says. “Companies will tend to spend their money in other places, like product development and customer research.”
Julie McCarthy is an associate professor of management at the University of Toronto. She says most large corporations now use some form of data analytics to inform their personnel decisions. Why? Because managers don’t have a very good track record of picking the best applicant. “The research shows us that human beings, although we want to trust our instincts, are not good at predicting performance,” she says, adding that relying on human intuition when it comes to hiring effectively amounts to rolling the dice. “But if we use these structured and standardized measures, we can be much more accurate.”
This is what McDonald’s Canada discovered when it began using its hiring tool. Jillard says he heard from several store managers who went ahead and hired employees despite being waved off by the software (store managers still exercise considerable discretion), only to find many of those same workers didn’t pan out. “Time and time again, they came back and said the tool was right,” Jillard says.
Data analytics is also touted as a way for employers to better manage their existing workers. Many firms now routinely conduct internal surveys that are designed to measure employee “engagement,” which studies have suggested can be a key indicator of a company’s future financial performance. Lowe’s, for example, discovered that high levels of employee engagement—say, asking a customer about their renovation project—drove an average store transaction four per cent higher. Added up, the difference between the stores with the highest employee engagement and the lowest amounted to more than US$1 million in sales a year, according to one industry report.
Survey data and statistical models can also reveal which employees may be likely to leave the company and suggest the best way to keep them. Depending on the circumstances, studies have shown that perks like extra days off and more flexible work hours are effective tools when it comes to maintaining employee job satisfaction. Data crunching can also help determine which employees should be targeted for promotion or are more likely to succeed in a different job. “Sometimes people get paranoid about this information being collected and held on them,” says McCarthy. “But if a company is simply holding information to find out what the best career progression is for this particular employee, or if an employee is getting stressed out or overwhelmed, that sort of stuff is good.”
Skeptics remain, however. Barsch of Teradata, for one, says the flurry of excitement around big data raises new risks. “It’s a tool, and a very valuable one, but I think there’s a danger in saying, ‘Well, the machine picked it so the machine must be right,’ ” he says. He points to the U.S. housing crash, which was driven in part by automated systems that were supposed to be able to determine who qualified for mortgages. “You can check their incomes and credit scores, but, as we learned, it’s also important to get a feel of who you are going to be lending money to,” Barsch says. “And I think it’s the same thing in HR. You want some analytics to help sort through the data, but at the end of the day you are hiring a person and that person, in most instances, is going to be talking to a customer.”
There are also opportunities for employees to game the system. “People are starting to get smarter by packing their resumés with key words,” says Barsch. The website Psychometric-Success.com sells an ebook for $14 that promises to tell you “how a potential employer ‘sees’ your personality on paper and how you can ensure your answers create the best possible impression.”
But as testing programs become more sophisticated, efforts to deceive them are likely to do more harm than good. “You’re better off just being honest,” says McCarthy. “Besides, it’s not just about the company selecting you, but you selecting the company. You don’t want to end up some place where the culture isn’t a good fit for you.”
Originally published in Maclean’s.