Software chooses the best candidate – Financieele Dagblad

Candidates interviewing is a waste of time. Who has sufficient historical data and the appropriate mathematical models, may consist of a stack of resumes her fine distilling who is best suited for a particular job. Indeed, as a recruitment specialist but sufficient data, it can predict how well someone will be in orbit without the person ever having seen

Sophisticated mathematical

For most companies, the above is a sketch distant future, but the technology there is already, scientist Colin Lee argues his thesis. He promoted this month at the Rotterdam School of Management (Erasmus University) research in which he uses a sophisticated mathematical model to analyze patterns in over 440,000 existing CVs and job applications. The model appears to predict with 70% accuracy who eventually actually invited for interview on the basis of such things as work experience, education and skills.

Intuition

” Important predictors of relevant work experience and years of service. You can merging into a formula, and thus determine the best match, “says Lee. Although work experience is crucial, recruiters are also not very consistent in what they allow you to decide, he concludes from the patterns. “We may as well recognize a common thread, but much like based on intuition to happen.

Suspicion

Where Dutch companies are hesitant to give a central role in the analysis of big data for recruitment and selection that practice has long been commonplace in Silicon Valley. Pioneers like Google base their hiring primarily on hard data and algorithms, based on successful recruitment of the past. “Companies are often extremely poor in interviewing and recruiting people. They sail by feel and baseless theories, “said Human Resources Director Laszlo Bock, Google last year in an interview with the FD.

can a company with pure data a path to the perfect candidate? In the Netherlands dominated the necessary suspicion, and not just about the still unproven technology. Ethical issues are involved, says Lee. “The future is that you can figure out exactly how someone is going to perform on the basis of the parameters in his resume. That’s scary because you people flattens in advance.

Optimal match

Recruiting Software is or has been used before in less extreme forms, such as by major broadcasters such as Randstad, Adecco and USG. Those with special software, a first pre-selection from hundreds or even thousands of resumes. This is done using so-called ‘applicant tracking systems (ATS). These are filters that both public information on social media and internal customer databases used to recruit, or to determine whether a worker is the best ‘match’ in his current post

‘Vaak we can better see if anyone comes into its own within a company than the company itself, “says Jan van Goch of Connexys, a maker of recruitment software. The main barrier to further development of these applications, according to him, not so much the technology, but rather the fear of privacy infringement and liability clients. They often sit on mountains of valuable historical information about their candidates, but refuse to unlock for use in larger databases.

Legislation

Van Goch: “If all this information comes together, we can do a lot smarter matching and recruitment . Customers want to do that, but they give do not always consent to use personal data and therefore remain seated and that’s a shame. A portion is afraid to be sued at the time that it comes up on the street, the more has been tightened since the legislation for data storage.