Big data. The new recruiting buzzword. As a matter of fact, its use is much less common than one might think. “So what gives?” with big data in recruitment, asks comedian Yvon Deschamps. A two-part series on the issue.
Big data is the collection of large volumes of data (gathered from social media, for instance) and exploited using algorithms.
Specifically, these models allow recruiters to predict the success and performance of an employee. It is a bit like the model that has long been used by insurance brokers to predict the degree of risk posed by an insured client. This risk is assessed on the basis of several data. Applied to human resources, it is called the recruitment data.
A recruiter could, for example, predict the availability degree of passive profiles, such as those found on LinkedIn that are not actively seeking employment, but that could be receptive to receiving offers. How would big data help in targeting these profiles? By analyzing the exact factors that influence this data. Thus, it was determined that those who regularly update their LinkedIn profiles or often change their pictures would be profiles that are more open to offers.
“These data should be looked at as tools available to recruiters in order for them to be more efficient. They do not, in any way, substitute the recruiter’s role,” states Eric Tonto, cofounder of Biglittlejob.com, a predictive recruitment platform.
When recruiting candidates for key managerial positions, data analyzes such factors as geographical distance, which has proved that the larger the distance between work and home, the higher the risk that an employee leaves within the first two years. “Having this information early on in the process allows you to adjust profiles, look for the right candidates based on specific features and review recruitment processes,” says Jean-Baptiste Audrerie, director of marketing and organizational psychologist at SPB Organizational Psychology. This type of data analysis was actually performed by Google, which has seen its entire interview process return to a more behavioural approach on the candidate's experience rather than clever questions aimed at destabilizing them.
Recent studies* reveal that recruitment would improve by as much as 25% when using algorithms rather than relying solely on the recruiter’s instinct. This percentage is based on incorrect recruitments, which can have significant financial impacts on a company.
Eric Tonto illustrates the impact of such tools on businesses. For a senior or strategic position, a flawed recruitment would cost around $ 300,000, if training, start-up costs (contracting) and reorganizational costs are taken into account. At 10 incorrect placements, this number approaches the two million dollar mark. An improvement of 10 to 20% of the recruitment process can make a real difference in business finances.
“Recruitment was previously an HR only platform. Scientific data will soon be joining this team. This is a real revolution!” states the Biglittlejob.com cofounder.
Still, the mission remains for big data to be rooted not only in language, but also in human resource practices.
* Study: http://hbr.org/2014/05/in-hiring-algorithms-beat-instinct/ar/1