The amount of data generated on and by individuals is rapidly surpassing conventional terminology—we’re entering the age of the “yottabyte.” While employers have yet to harness and analyze this data, a new industry is developing to do just that. The discussion during the final session of yesterday’s conference on the future of work brought together entrepreneurs at the forefront of this industry to share their views on how employers can, and should, use data to optimize their hiring practices, their workforce, and ultimately, their business.
Seyfarth Shaw LLP hosted Work: The Future, a gathering of business and industry leaders focusing on the future of work for employers, on June 21, 2017 in San Francisco. The panel was moderated by Seyfarth partner and lead of the firm’s nationwide workplace counseling and litigation group, David Baffa, and included Max Drucker, CEO of Social Intelligence at Carpe Data, and Loren Larsen, CTO of HireVue.
Baffa began the discussion by asking how HireVue and Carpe Data see data being utilized by employers. He introduced Larsen, who discussed HireVue’s mission and process. On the employer side, HireVue strives to help companies find the best talent to fuel growth. On the employee side, HireVue aims to provide candidates with a platform to tell their unique story through video. As to specifics, HireVue is essentially an intermediary in the hiring process, allowing employers to create question guides, which candidates respond to in video form. Once the candidates’ videos are collected, hiring managers have fingertip access to all of the video interviews and can select candidates efficiently and effectively, comparing responses to specific questions and choosing the best candidates. In addition, HireVue offers the additional service of working with its clients to develop a validated assessment process by taking in performance data, and developing an algorithm to analyze interview data across candidates to ultimately provide the company’s HR team with a pool of the most ideal candidates for a specific job based on demonstrated correlations between interview attributes and successful job performance. The speed of the process also provides applicants with an exceptional candidate experience. It’s a win-win.
Drucker’s Social Intelligence focuses more on employers, providing a streamlined process to conduct pre-employment social media background checks. Drucker noted that employers, and particularly hiring managers, often avoid reviewing applicants’ internet profiles for fear of treading on fair and equal employment practiced. In doing so, however, they risk exposing their companies and other employees to negligent hiring practices. He said that Social Intelligence helps mediate the risks on both sides. It does so by limiting the breadth of its background searches to within legally acceptable parameters. Specifically, Social Intelligence conducts background searches on candidates in the following areas only: including past illegal or violent conduct, explicit materials, and demonstrations of racism and/or intolerance. Ultimately, Drucker warned that ignoring applicants’ social media and internet presence is plainly reckless.
Potential Risks for HR: Discrimination? Privacy?
After hearing about these cutting-edge technologies, an audience member chimed in with the question of the day: while the algorithms and processes of HireVue and Social Intelligence sound great, how do companies avoid stepping into adverse impact territory? Larsen said that algorithms can be tuned and tweaked in real-time to ensure that the candidate pool being generated by HireVue does not turn out to adversary impact any particular demographic. Drucker added that the parameters around use of HR data also depends heavily on the situation and position involved, giving the example of sales employees, whose progress or success is more easily quantifiable than, for example, a benefits associate.
Larsen also pointed out that data can actually be used to promote diversity and limit some biases inherent to human functions. As to HireVue, Larsen said that its algorithms could be tuned to create a more diverse workforce through the sourcing of more diverse candidates—something that would benefit employees and the bottom line.
Privacy is also an issue to consider in using data on potential (or current) employees. In responding to a question about the privacy implications of using personal data, whether through an interview platform or social media, Baffa highlighted what he views to be the clearly unacceptable forms of data analysis: using data on candidates or employees to select protected categories or discriminate on a health and wellness basis.
Baffa concluded the session with some words of wisdom for employers looking to bring their HR systems into line with technological advances in data processing and usage. In general, he advised that companies should aim to use data in smart ways and for clear business purposes. He also offered the following best practices:
- Define what you are trying to learn, fix or solve, and then evaluate potential sources of data relevant to those goals.
- Think about whether adverse employment consequences are likely to flow from the data processes you are trying to build.
- Work with vendors who understand the risks involved and can help mitigate those risks, including the development of validation studies and avoidance of adverse impact.
- Pay special attention to disability discrimination, and avoid leveraging data on health or wellness, or trying to predict characteristics such as absenteeism or likelihood of workplace injury.
- Make sure any data correlations you use are logical overall and justified from a business perspective.
- Keep in mind that data insights should inform and enhance your judgment and decision-making, not own it.