Artificial intelligence helps Japanese companies seeking to reduce staff turnover

April in Japan coincides with the first days at work for new employees fresh out of university. These young people are valuable to companies facing an acute labor shortage. However, one in ten young people leave their job within a year of being hired – sometimes from the first day – and 30% do so within three years, according to the Japanese Ministry of Labor.

Hence the idea of ​​Naruhiko Shiratori, professor at Tokyo City University, to develop a system based on artificial intelligence (AI) assessing the probability of an employee leaving, in order to prevent it. Developed with the start-up Shikinami, the system compiles data relating to the company’s employees, such as their attendance, their age and gender, and their leave. It also includes information concerning employees who have left the company. This helps create a company-specific employee turnover model.

Based on this, the user can enter data about new recruits, such as their arrival and departure times and any absences after joining the company. This makes it possible to assess the risk of resignation. When tested in several companies, the system’s accuracy rate reached 83.7%.

Impact of discomfort

To create this tool, Mr. Shiratori relied on software that he himself developed to limit dropping out of studies. “On average, the dropout rate in universities is less than 2%, but in some establishments it can reach several tens of percent. » This has a significant impact in particular on the finances of private establishments, themselves faced with the general decline in the number of students.

Read also | Article reserved for our subscribers Japan relies on new technologies to overcome labor shortage

The challenge is similar for companies wishing to reduce costs – recruitment, training, vacancies and retirements – generated by personnel changes. A hire costs on average 936,000 yen (5,600 euros) for new graduates and 1,033,000 yen (6,200 euros) for mid-career workers, according to the 2020 edition of the White Paper on Employment of the Recruit classified ads specialist.

Read the column: Article reserved for our subscribers Managing with artificial intelligence: recruiter biases

At the same time, the costs linked to the psychological impact of an employee’s discomfort are estimated at 4.22 million yen (25,000 euros) by the Government Council for Equality between Women and Men. “The mental health of employees and the reduction of staff turnover, such as absences due for example to harassment problems, are important issues for the sustainability and competitiveness of the organization”underlines the start-up Shikinami on its site.

You have 29.28% of this article left to read. The rest is reserved for subscribers.

source site-29