Gut feeling or AI — which one can sense when an employee is ready to quit?

Maintaining a stable workforce and retaining valuable talent is not easy. To succeed in doing so, it is important for organisations to identify the employees who are likely to quit. While some managers may rely on their gut feel or intuition to sniff out employees who may be contemplating leaving, artificial intelligence or AI-based data analysis can provide a more objective and data-driven approach.

Both approaches have their own strengths. So then, which one is the more accurate of the two?

“AI-based tools are helpful in analysing the data to identify patterns in employee attrition,” says Manish Majumdar, head-HR, EMS, Centum Electronics. These tools use AI algorithms to determine when employees are most likely to leave, such as after a period of 24 or 36 months of employment. By identifying these patterns, organisations can focus their efforts on retaining employees who are approaching these milestones, as their likelihood of quitting or moving on may be higher. For instance, if most employees leave after 36 months, organisations can start focusing on retaining employees as they approach their 20th or 30th month of employment.

“AI facilitates the process by accessing the huge volume of data that even the companies sometimes don’t have bandwidth for,” opines Rishav Dev, CHRO, Noveltech Feeds.

“While I recognise the benefits of using AI tools to analyse data, I do have concerns about the parameters being used and the quality of data input.”

Manish Majumdar, head-HR, EMS, Centum Electronics

Additionally, “AI-based data analysis can provide a more objective and data-driven approach to identifying patterns and trends related to employee turnover,” says Mukul Harish Chopra, CHRO, ConveGenius.  By analysing large amounts of data, AI algorithms can identify factors that may be contributing to turnover, such as job satisfaction, work-life balance, or compensation.

However, can data beat the ingenious instincts of human gut feeling?

Not really, according to Chopra who explains this by citing the example of the recent programme launched by the Indian government to evacuate people in Sudan. “Under ‘Project Kaweri’, Indian pilots are evacuating people using the C-130 Hercules, which is an all-weather, all-terrain aircraft capable of landing in extreme conditions. However, the aircraft was not programmed to land on the airstrip they are using, as certain conditions needed to be fulfilled. In response, the pilots used their ingenuity and improvised. They used night-vision glasses to conduct a landing that their aircraft wasn’t even designed for. While night-vision glasses are typically used by special forces, the pilots were able to use them successfully to land the aircraft, even though it wasn’t meant for it. This situation demonstrates the importance of human ingenuity and problem-solving skills, even in the presence of advanced technology.”

“Humans have a unique ability to behave in seemingly contradictory ways under similar circumstances, making them the most intelligent beings. That is why, AI should be viewed as an aid to human intelligence, rather than a replacement for it.”

Mukul Harish Chopra, CHRO, ConveGenius

He adds, “Humans have a unique ability to behave in seemingly contradictory ways under similar circumstances, making them the most intelligent beings. That is why, AI should be viewed as an aid to human intelligence, rather than a replacement for it.”

“The role of AI in natural language processing is limited to what it has been trained to do. It only understands what we teach it and operates based on scientific principles. If the data goes wrong, the judgement will also be wrong,” points out Dev. AI-based data analysis may also miss important contextual factors that are not captured by the available data.

Some AI software may consider factors such as decreased productivity rates while analysing data to identify potential quitters. However, the success of AI algorithms relies heavily on having the right data sources and parameters. “While I recognise the benefits of using AI tools to analyse data, I do have concerns about the parameters being used and the quality of data input,” points out Majumdar.

“The role of AI in natural language processing is limited to what it has been trained to do. It only understands what we teach it and operates based on scientific principles. If the data goes wrong, the judgement will also be wrong.”

Rishav Dev, CHRO, Noveltech Feeds

For instance, AI software can be used to analyse various data points such as age group, gender, location, department, designations, HODs, managers and performance records to identify the factors contributing to employee attrition. This can be especially useful for larger organisations with a large amount of data to crunch. However, it is important to note that human perception and gut feeling are still important for managers, as there are certain nuances and factors that AI may not be able to pick up.

“The process of quitting is not spontaneous but accumulated over time, and machines may only track 90 per cent of the necessary information. While technology can help manage a large team, it still requires the human touch to truly understand and connect with employees,” believes Chopra. Technology cannot help retain disengaged employees if there is no human connection to begin with. Building relationships and listening to employees’ concerns is crucial in retaining them.

In managing a team, it is crucial to have a finger on the pulse of the people, regardless of age, gender, location or other parameters. Even dedicated and hardworking employees may want to leave for various reasons. “While AI can be useful to analyse data, it cannot replace the importance of human intuition and connection. A manager can pick up cues and sense moods that AI may miss, such as a remark or a reaction to something. Data analysis can sometimes be inaccurate, but people’s behaviour can reveal their true feelings about their job,” believes Majumdar.

Majumdar recalls a personal experience in his organisation where a person was supposed to join a meeting but didn’t show up. This was unusual, and upon investigating, it was discovered that the person, unhappy with their portfolio, had been contemplating changing it, and seemed willing to leave, if there was no solution in sight.  “This is not something that AI would necessarily sense as it doesn’t fit into the usual data points such as tenure, age, gender and so on,” observes Majumdar. However, he feels, “It’s important for managers to be perceptive and mindful of how their employees are feeling, even if it doesn’t show up in the data”.

Agreeing with the statement, Chopra also shares, “Human gut feeling, or intuition, can sometimes pick up on subtle cues that may not be immediately obvious from the data.” For instance, managers who have a good relationship with their teams may be more likely to notice changes in behaviour or mood that may indicate that an employee is considering quitting.

If an employee is considering switching, their immediate supervisor or manager would be the first to know. “Therefore, every team leader or manager should have a good relationship with their employees and show empathy towards them, so that they can detect any signs of disengagement early on. This is the most effective way to prevent employees from leaving,” opines Dev.

Data crunching can be helpful in facilitating the process, especially in big organisations, but one cannot rely solely on AI.

“Data is meant to assist the process and make it more efficient, but it cannot be relied upon solely. There is a possibility of missing important information as emotions cannot be entirely captured in data. For instance, if someone is dissatisfied with their job, they may not openly express it, and may only confide in their close friends. In situations where the human element is lacking, AI can provide some valuable insights, but it cannot replace the need for human relationships and connections,” concludes Dev.