Enlisting Analytics To Level The Hiring Landscape

Finding those candidates is particularly challenging right now: In April, the U.S. unemployment rate dropped to 3.6%, the lowest figure since 1969, according to the U.S. Bureau of Labor and Statistics. And the squeeze is particularly acute in fields where the talent pool is shallow, such as data science, and in fields that are overtapped, such as software development. “As a company and as an industry, the biggest challenge we face globally is the shortage of talent,” says Brad Cook, Teradata’s global vice president for talent acquisition. 

So how can companies compete effectively in this tight talent market? Below, leaders at three innovative organizations explain how they’re giving recruitment a data-driven reboot, leveraging analytics to solve three hiring challenges: bias, inefficiency and time. 

Hiring continues to be stacked against women and minorities. Recent studies show that minorities are twice as likely to be invited to interview for jobs if they “whiten” their resumes and that levels of hiring discrimination against African-Americans have not improved in 25 years. Women are persistently hired less than equally qualified men, in part because, according to researchers, hiring panels fundamentally believe that men are more skilled at performing certain professional tasks. And yet, diversity is a proven benefit to organizations, not just culturally but financially: companies in the top quartile for gender and ethnic diversity beat their competitors in performance by 15% and 35%, respectively. How can companies shake gender and racial bias, particularly when that bias shows up in hidden ways? 

Ten years ago, algorithms were proposed as the solution; if humans are bigoted, let data make the decisions. But algorithms are trained on data provided by humans and thus subject to bias, too. 

After a decade of watching such prejudices in action, neuroscientist Frida Polli, CEO and cofounder of Pymetrics, set out to design a recruiting process free from bias. She knew that would be impossible to accomplish if the data used to evaluate candidates hinted at traits like gender or ethnicity. Resumes, for example, tend to trigger all sorts of implicit prejudices; even if names are scrubbed, it’s still obvious that an applicant who played football is male. 

Pymetrics came up with an unconventional solution: Gamify hiring. Its software screens candidates through a series of neuroscience games that identify cognitive, social and emotional traits — none of which are correlated to gender, race or age. Highly successful existing employees at the hiring company play the same computer games, so Pymetrics can identify the traits of overperformers in each role. These might include skills like memory, altruism or risk aversion, which are among the 90 traits tested for. 

Then an algorithm compares candidates’ and existing employees’ scores. “We say, ‘Okay, what makes people in finance different from people in HR or marketing?’ and get to a specific algorithm that’s unique for each role in that company,” says Polli. Sure enough, the candidates whose results most closely match the traits of ideal employees are diverse by all measures that typically face bias (age, gender, ethnicity, background, etc.). On average, according to Pymetrics, the process saves recruiters 250 hours, leads to twice as many hires and boosts gender, ethnic and socioeconomic diversity by a minimum of 20%.

Hiring the right person can feel like finding a needle in a haystack, and hiring the wrong person is expensive. Yet, while AI has proven to be a promising professional matchmaker, just 4% of executives say their HR teams are advanced users of analytics today, according to a survey by Forbes Insights and Teradata of 600 senior executives and data scientists in Q2 2019.

LinkedIn wants to bring AI-enabled hiring to the remaining 96%. Three types of data help the company pair candidates and recruiters, says John Jersin, vice president of product for LinkedIn Talent Solutions: “explicit data”, information users enter into their profiles or job search boxes; “behavioral data,” such as which recruiters message users and when; and “inferred data,” which reads between the lines, gleaning that, say, a user neglected to enter skills they obviously picked up when they left a role in fashion to become a UX designer, and allowing LinkedIn to show them more, and more relevant, opportunities. 

When Microsoft bought LinkedIn for $26 billion in 2016, the company decided to expand on those services, harnessing what was then a 450 million-user-strong data cache to enter the business intelligence market. In addition to matchmaking individual recruiters and candidates, it began leveraging analytics to help organizations track macro-trends around talent supply and demand,such as where workers with specific skills are located. Now, with a few clicks, companies can figure out how big the sales workforce is in, say, New York versus San Francisco, which enterprises are using to determine “where they should place their bets strategically” and drive team building, says Jersin. 

Another critical but often overlooked decision LinkedIn helps companies make: how to describe positions. “We can show you how the market data fits different definitions for that role,” Jersin says, noting that a job description “has a cascading influence across who you're going to search for, who you're going to message and who will interview.” 

The hope is to truncate the tedious search process, enabling recruiters to focus on what they do best: “Hiring requires a recruiter to sit down with a candidate and understand who they are and whether they’re a cultural fit for that company,” says Jersin. “I think we’ll see an increased investment in the human process as recruiters spend less time wading through data and profiles.”

Laura Nyquist, Teradata’s chief human resources officer, leads a department that hires for 2,000 positions a year. Most are in tremendously competitive markets “with low-low unemployment,” she says. For tech talent, which makes up a large part of the cloud analytics software company, unemployment has plummeted to a mere 1.3% and reasonable time to hire an employee hovers around 50 days — which means that while one company convenes an interview panel for the perfect candidate, another company might just hire her. 

To better compete in this high-velocity environment, Teradata studies its own recruitment process, driving speed by tracking speed.

Brad Cook, Teradata’s global vice president for talent acquisition, breaks each stage of the application-to-hire process into a quantifiable segment, and day-to-day figures provide instantaneous insight into lags. The objective is shorter spans, leading to successful hires. Otherwise, recruiting costs rise steeply, and candidates are scooped up by competitors. 

Time is a valuable metric in the hiring process, not just because it’s in short supply but also because it can help pinpoint all types of inefficiencies and bottlenecks in recruitment — and signal teams when to make a change. Cook can monitor, say, Teradata’s average interview-to-hire time in Asia or how many candidates have been contacted for an open position. If 100 is the standard and he sees only 43 in the system, he can swiftly prevent a catastrophe. If the click-through rate on a job ad is high but few people apply, perhaps the talent supply has suddenly changed in that city or the job description isn’t clear; knowing that, he can relocate or redefine the position. 

The end goal is efficiency.“When we drive efficiency, that pushes down our cost and speeds up our time in filling positions,” says Cook, who adjusts hiring processes in real time based on the trends these analytics reveal. “It’s what helps us make decisions on what to do next.”

Source: Forbes.com

Powered by NewsAPI.org

Keywords:

Unemployment in the United States • Unemployment • Labour economics • Statistics • Angle • Discipline (academia) • Skill • Data science • Software engineering • Company • Industry • Teradata • Globalization • Vice president • Recruitment • Company • Market (economics) • Leadership • Innovation • Organization • Recruitment • The New 52 • Recruitment • Bias • Recruitment • Minority group • Minority group • Employment • Résumé • Discrimination • African Americans • Human trafficking • Professional certification • Employee benefits • Organization • Company • Quartile • Gender • Multiculturalism • Performance • Gender • Racism • Algorithm • Human • Decision-making • Algorithm • Data • Human • Subject (philosophy) • Bias • Prejudice • Action (philosophy) • Neuroscience • Scientific method • Political freedom • Bias • Gender • Ethnic group • Football • Neuroscience • Game • Cognition • Society • Emotion • Trait theory • Race (human categorization) • Ageing • Employment • Recruitment • Company • Play (activity) • Skill • Memory • Altruism • Risk aversion • Algorithm • Employment • Finance • Human resources • Marketing • Algorithm • Trait theory • Idealism • Employment • Multiculturalism • Bias • Ageing • Gender • Ethnic group • Ethnic group • Scientific method • Gender • Socioeconomics • Artificial intelligence • Professional wrestling authority figures • Human resources • Data analysis • Survey methodology • Forbes • Teradata • Senior management • Data science • LinkedIn • Artificial intelligence • Jersín • Vice president • LinkedIn • Data • Information • Behavior • Data • Data • Skill • A/UX • User experience design • LinkedIn • Microsoft • LinkedIn • Company • Service (economics) • Consumer • CPU cache • Business intelligence • Market (economics) • Data analysis • Organization • Macroeconomics • Skill • Supply and demand • Workforce • Skill • Company • Workforce • San Francisco • Business • Team building • Jersín • Decision-making • LinkedIn • Jersín • Cognition • Recruitment • Jersín • Business process • Data • Teradata • Chief human resources officer • Competition (economics) • Unemployment • Cloud analytics • Unemployment • Employment • Company • Teradata • Recruitment • Business process • Teradata • Multinational corporation • Vice president • Recruitment • Insight • Recruitment • Shortage • Recruitment • Teradata • Click-through rate • Advertising • Person • Skill • Goal • Efficiency • Motivation • Efficiency • Business process • Decision-making •