Can an AI predict when you will leave your job? Yes, and that's how it works
Are you about to quit your job? Chances are your bosses already know about it. During a report published in 2019, CNBC commented on a new technology that allowed companies to know when their workers were close to leaving the ship. In fact, the artificial intelligence of the IBM company is capable of predicting with 95% certainty who are the workers who plan to resign from their position.
This was confirmed by Ginni Rometty, who was president and CEO of the company until early 2020. In the aforementioned CNBC report dating from 2019, Rometty talks about the technology used. In fact, she comments that they have been using artificial intelligence to retain the company's talent . “The best time to approach an employee is before she leaves,” the former CEO commented at the time of her.
This artificial intelligence is so widely used within IBM that it even has its own patent. In this way, your human resources department can use the “leave prediction program” to control your employee population. Thus, they can act before it happens and arrange actions to create a rapprochement between managers and the vulnerable group.
Of course, Ginni Rometty has not revealed the secret behind this AI. However, the former CEO commented that the reliability of this technology had to do with the analysis of a lot of data within the company . An artificial intelligence almost as impressive as the one that guesses your gender just by looking into your eyes.
A most interesting algorithm
Photo by Florian Olivo on Unsplash Professor Brooks Holtom of Georgetown University published a little more information on IBM's artificial intelligence in the Harvard Business Review. According to him, these predictions can be made by using computer learning technologies and large amounts of data together. Thus, the necessary information can be measured to know if an employee is about to resign.
Among these types of data, Holtom describes some. In the first place we have events that can cause the departure of personnel. These can be describedas changes in company leadership or a major purchase by another larger company. In addition, other factors such as news related to this type of event, changes in the stock market and legal actions against the entity were also taken into account.
In addition, we have another factor: the connection of the workers with the company in which they work is essential for this process. Using public data such as number of previous trades, employment anniversaries, skills, education, gender and geographic location, some conclusions can also be drawn.
The benefits of an AI how are you
Although many may take this movement as an invasion of privacy —a topic that invites debate—; this type of artificial intelligence could end up being beneficial for employees. In fact, many companies prefer to financially compensate dissatisfied workers, instead of simply replacing them with a new person.
The cost of replacing a highly skilled worker could end up being twice their salary. For this reason, instead of wasting time and money on selecting, hiring and training a new professional, some companies focus on making their current workers happier.
Even if you can predict who is going to leave, you still need to respond thoughtfully.
Brooks Holtom, Professor at Georgetown University Holtom describes in his article that “the best organizations develop a culture of listening.” Thus, managers can hear from their employees what things are going well, and which are not so good. In this way, working conditions within the company can be improved.
With this artificial intelligence technology, managers and workers can create a better map of what each one really values. They might also be able to develop a more positive environment, and a personalized work style. Of course, the concern about privacy prevails:
My biggest concerns are regarding privacy limits. At a certain level, these could even become invasive. Pulling information from Instagram, Snapchat or Facebook. Information that people upload publicly, whether password controlled or not. There is a risk for employees.
Brooks Holtom, professor at Georgetown University