You are interviewing a candidate for a position at a call center. You need someone polite, courteous, patient, and dependable. The candidate you are talking to seems nice. But how do you know who is the right person for the job? Will the job candidate like the job or get bored? Will they steal from the company or be fired for misconduct? Don't you wish you knew before hiring? Retail employers do a lot of hiring, given their growth and high turnover rate. According to one estimate, replacing an employee who leaves in retail costs companies around $4,000. High turnover also endangers customer service. Therefore, retail employers have an incentive to screen people carefully so that they hire people with the best chance of being successful and happy on the job. One company approaches this problem scientifically, saving company's time and money on hiring hourly wage employees. Evolve finds data-driven predictors of job performance and uses this information to help select the right fit for the job. In October 2014, Cornerstone OnDemand, a publicly traded talent management company, acquired Evolv for $42.5 million, potentially extending its reach.
The idea behind the software is simple: If you have a lot of employees and keep track of your data over time, you have access to an enormous resource. By analyzing data from a large number of employees, you can specify the profile of the "ideal" employee. The software captures the profile of high performers, and applicants are screened to assess their fit with this particular profile. As the database gets larger, the software does a better job of identifying the right people for the job. Employers such as Xerox are using the software developed by Evolv, where job applicants complete a test that takes half an hour. The system compares the applicant to the ideal profile, and the hiring manager gets a color-coded message from the system, where green indicates a high potential employee. Xerox won't even look at a resume if the system generates a red sign.
The profile of the ideal candidate is often counterintuitive. For example, data on call center employees indicate that the best candidate has a short commute to work and participates in a small number of social networking sites. Contrary to what some people may think, job-hopping and unemployment status are not good predictors of effectiveness in the next job. One thing the system pays a lot of attention to his personality. It seems that for call center workers, being inquisitive results in leaving the job sooner. The system also measures honesty. For example, one question asks candidates to report how much computer skills they have, and then a follow-up question asks what control-V does.
The users of the system praise the time-savings and the results: Xerox saw increases in performance and reductions in the turnover of their employees after adopting the system. On the negative side, anti-discrimination lawyers think that this is new territory with potential legal downsides. Moreover, these systems are used only for hourly or retail workers where data exists for thousands of employees and the system can identify a reliable employee profile. Its applicability to higher-level, professional, and more unique jobs is not yet clear. How big data approaches change the face of selection continues to evolve, including becoming Cornerstone OnDemand.
This case was written by Berrin Erdogan and Talya Bauer to accompany Bauer, T., & Erdogan, B. (2015). Organizational Behavior (2nd Edition). Washington, DC: Flat World Knowledge. Based on information from Fastenberg, D. (April 10, 2013). "Big data" predicts who makes the best workers. Retrieved June 2, 2014 from jobs.aol.com/articles/2013/04/10/big-data-evolv-hiring-employers/; Ito, A. (Oct 24, 2013).
Hiring in the age of big data. BusinessWeek; Leber, J. (May 27, 2013). The machine-readable workforce. MIT Technology Review; Lohr, S. (April 20, 2013). Big data, trying to build better workers. New York Times.
Questions
1. Have you ever taken part in a selection system such as the one described in the case? How do you feel about these tests as a job candidate?
2. Do you think the use of big data to incorporate information about an individual's personality is a useful or problematic approach? Justify your answer.
3. Should organizations care about how job applicants react to pre-employment selection tests?