Advertisement
A Growing Question: Just How Effective is Big Data in Hiring?
Get articles like this
in your inbox
The longest running and most trusted source of information serving talent acquisition professionals.
Advertisement
Big data is transforming the information world at an alarming rate. It’s no surprise that the data being collected, sliced, sorted, and sold is being used to help businesses make better hiring decisions.
According to The New York Times, a growing number of entrepreneurs are applying big data to human resources and the search for talent, creating a field called workforce science. To quote technology writer Steve Lohr, “workforce science, in short, is what happens when big data meets HR.”
The data part of the equation is collecting all of the digital imprints that a worker or job seeker has left in the course of web browsing, emailing, instant messaging, or posting on social media. The “science” part — evaluating and measuring the data — is the tricky part. Silicon Valley start-ups are developing analytics and algorithms to put this data to work for recruiters and employers.
One way that big data is being tapped for HR is the internal collection of information on the job.
Companies are collecting data on current employees through existing ATS and HRIS systems, tracking mechanisms and software. Employees are surveyed, tested, and monitored. That information is then analyzed throughout the employee’s tenure, and certain traits and behaviors are identified. The company can determine patterns in characteristics and behaviors that lead to a successful and engaged employee, and use that information to recruit for similar qualities in new hires and to steer the right people into the right jobs.
One example of this approach is IBM. A couple of years ago, IBM acquired Kenexa, a software company that focuses on recruiting, hiring, and training. At the time of the acquisition, Kenexa was surveying and assessing over 40 million job applicants and employees per year. Using the Kenexa technology, IBM has branched out and is now helping other companies mine their own “big” data by providing the platform to tap into their own gold mine of information.
A number of emerging companies are jumping into the water. Start-ups are tapping into the vast sea of public information available from prospective job candidates and channeling the results into software tailored for hiring managers and employers.
One such company is hiQ, which claims to focus on “people analytics.” It gathers public data, a business’ internal data, or both, to create predictive models.
Evolv, another early innovator that used big data for HR, was named to Fast Company’s list of the World’s 10 Most Innovation Companies in big data in 2014. Evolv, now part of Cornerstone software, developed a social sourcing model to assess job candidates and employees by looking at more than 500 million data points on things like gas prices, unemployment rates, criminal backgrounds and social media usage.
Knack, a Silicon Valley start-up firm, uses a different approach. It has developed a series of computer games that test emotional intelligence, cognitive skills, memory and risk-taking.
Gild is another company that is harnessing big data for employers. The company made a big splash in the HR world with software that helps recruiters find qualified developers and other technology talent. Gild’s model is to analyze candidates’ publicly available code, using proprietary algorithms to analyze their technical ability gathered from open source communities and Q&A sites. It also uses job and academic history to profile and score the developers, and then ranks them by ability to code, and their competitive rank.
So far this all sounds good, right? But critics say there can be negatives.
Much of the data collection done by companies like Gild is happening without transparency and consent. Concerns about privacy and consumer protections are developing as quickly as the new technology, and have made their way to the White House. The President issued a report on privacy and big data last year, and rolled out new initiatives this year to regulate information and protect consumer privacy.
Data collection on the job has privacy advocates crying foul. Many are alarmed at the notion that Big Brother is alive and well in the American workplace.
Proponents of big data claim that the information collected is neutral and without human bias. The counter-argument is that the data collected contains inherent prejudices and it perpetuates larger societal bias.
The concern is that using information mined from a variety of public sources could result in reinforcing pre-existing patterns of inequality and exclusion. If so, the unintended consequence is that the use of algorithms and data crunching would have a disparate impact on minorities or other protected classes.
Researchers Solon Barocas and Andrew Selbst have published a fascinating study exploring this theory through the Social Science Research Network.
If a disparate impact in the use of big data could be proven, agencies like the EEOC would likely engage in preventing any potential discrimination in the workplace. And that’s a chilling prospect for most employers. Can you say “individualized assessment?” Enough said.
The information collected and used by data brokers and software companies may be in the public domain, but does that make it fair game? The FTC has stepped in on more than one occasion to crack down on web sites collecting big data for hiring purposes. The commission fined data broker Spokeo $800,000 for collecting publicly available information and repackaging it for recruiters and employers.
The argument used against Spokeo — that is was violating the Fair Credit Reporting Act (FCRA) by proving reports without affording the required consumer protections — could extend to other users of big data. This is particularly true when the information is being used to make hiring decisions about potential candidates and existing employees without any of the mandatory safeguards like disclosures, authorizations, and adverse action notices required under traditional background screening methods. Employees and job seekers typically have no way of knowing what information is being collected and how it is being used.
LinkedIn is currently fighting a similar case in a California court. The plaintiffs are claiming that LinkedIn has gotten into the background screening business, but isn’t playing by the rules.
The argument goes like this: Since the information on LinkedIn is information compiled and sold for use in employment, all of the protections and requirements of the FCRA should apply. The same argument could be made in cases where companies use algorithms to filter big data and rank candidates.
At the end of the day, the jury is still out as to whether big data tools are effective in hiring. The success of programmatic tools for measuring certain technical skills can be invaluable.
But ultimately, when making a hiring decision, employers need to know whether a candidate can do the job. All of the indirect assessments and scoring in the world cannot predict success on the job as effectively as human observation and direct evaluation of a candidate.
Given all of the potential benefits, legal risks, and variables in this developing science, employers should consider the following tips:
As always, consult your legal counsel to review your recruiting and hiring practices and to ensure that you are engaging in best practices to protect your company as well as your employees and job candidates.
This was originally published on the EmployeeScreen IQ blog. EmployeeScreen IQ is not a law firm, and the contents of this article are not intended to be a substitute for legal advice.