Predictive Recruitment Discovers the Best Employees through Algorithms

Predictive Recruitment Discovers the Best Employees through Algorithms

FacebooktwitterlinkedinmailFacebooktwitterlinkedinmail

Currently present in finance and marketing, algorithms are starting to win in the recruitment sector.

BHV Marais found the solution for discovering the rare pearl and avoiding the errors of hiring: predictive recruitment. “This is a technology that, through algorithms, reliably predicts the likelihood of success of an employee. By success, we mean performance and doing it to be happy and engaged in your work” explains David Bernard, manager of AssessFirst, a business specialized in the concept of of predictive recruitment platforms.

“The utilization of predictive models based on algorithms has existed for around 20 years in finance or in marketing for predicting consumer behavior. Since 2014, this type of model started to appeal to the world of recruitment because it reduces uncertainty”, estimated David Bernard. The offering of predictive recruitment by AssessFirst has been around for 18 months and has already produced 18,000 recruitments.

Today more and more businesses are entering the market, from giants of human resources like Cornerstone to companies specialized in psychometric tests such as CEB or the Quebec-headquartered company D Teak.

“The powerful rise of predictive recruitment is linked to Big Data becoming increasingly reliable. Every day I also observe that the HR function is ready to use more and more procedures to reduce risks”, explains Jean-Baptiste Audrerie, organizational psychologist specializing in recruitment algorithms.

The foundation of a recruitment platform is established in the following way: “First, the designer of the platform can do an internal audit in the business to measure the skills that are necessary to be successful in the valued position. For example, it can be based on the evaluation of professional interviews or job descriptions. From there, the algorithm is developed. However, the majority of predictive recruitment platforms possess general models that measure reasoning ability, motivation, and the necessary personality to be successful in certain positions, ranging from commercial to advising clientele. The selections of positions is vast”, said David Bernard.

Once the algorithm is ready, the specialists design a questionnaire that includes between 50 and 100 questions to which the candidate responds to in about 15 minutes. It could ask questions about an individual’s perfectionism, selling ability, or a relationship with a client. At the end of the questionnaire, the employer receives a percentage of correspondence between the candidate’s profile and the expectations of the business. “Everything depends on the expectations of the recruiter but we think that with a correspondence match of 60%, it is appropriate to call the candidate for an interview”, explained David Bernard.

Benefits for Businesses

Fathallah Cheref, director of human resources at BHV Marais, is a supporter of predictive recruitment. For a year, he has used the the platform developed by AssessFirst. “Each year, we recruit between 500 and 600 employees. But our recruitment process was very time consuming. We started with a phone interview, a group interview with 3 or 4 candidates. Despite this, the casting errors were too high, we had to change something about it.”

To avoid a high employee turnover and the long process of recruitment, Fathallah Cheref and his team developed with AssessFirst a “job profiler”, which is a custom-designed algorithm.

The goal was to identify and map out the necessary skills to be a good seller, measuring qualities such as customer service or selling ability in 15 minutes”, explained the director of human resources.

Finally, the results are met: “With the algorithm, we reduced the recruitment process by a third of the time, which went from 45 to 30 days while the turnover rate was significantly reduced. Now we also use predictive recruitment for our managers”, said Chereft.

… and for employees

“The development of predictive recruitment provides a legitimate fear: the risk of dehumanizing recruitment by leaving the algorithms alone to select the candidates without the difficulty of meeting. But by analyzing good results, we realize that this is not the case at all. On the contrary, this method of recruitment is profoundly humane to the extent that it reveals the talents of everyone regardless of school, type of degree, origin, or age”, explained Jean-Baptiste Audrerie.

This thought is shared by David Bernard: “Predictive recruitment can reveal potential that recruiters would never see in an interview. A real example of this was Berner, a company that sells professional equipment, that recruited a salesperson who beat all of the sales records. It was a florist who knew nothing about the sector. With a C.V. and a traditional cover letter, she never would have received an interview”, the specialist recounted.

Furthermore, predictive recruitment allows applicants who land an interview to have a constructive exchange. “From the start, the recruiter knows that the candidate is potentially reliable. There is no need to waste time asking about degrees, qualities, or flaws”, explained David Bernard.

However, applicants who are not selected for an interview could find it unfair to be eliminated after a simple 15 minute test. “This can be disconcerting”, admits David Bernard. “But the candidates that are not selected for an interview receive personalized feedback that explains what industry, position, or company size corresponds the best with their profile. In nine months we will even be able to offer information that corresponds with the vacancies of the platform. This is more fair than a robotic message announcing to a candidate that they did not receive the job for which they interviewed for even though they thought they were successful in that interview.”

FacebooktwitterlinkedinmailFacebooktwitterlinkedinmail
Brendan Dougherty

Leave a Reply

Your email address will not be published. Required fields are marked *