Published on the 21/08/2019 | Written by Jonathan Cotton

It’s the next step in digital HR management – AI-powered employee flight-risk prediction. And yes, it’s available to regular schmoes too…

Ask any HR manager and they’ll tell you: Replacing employees is a hugely expensive, time consuming process.

But what if there was a better way? There is, says IBM. Using Watson, AI and its HR-centric suite of employee management products, HR managers can now predict which employees are likely to hand in their resignations – and receive advice on what to do about it.

That’s the takeaway from a recent interview with IBM CEO Ginni Rometty who says that, thanks to IBM’s Watson-developed ‘predictive attrition program’, the company can now predict which employees would leave employment in the next six months without intervention and can do so within a 95 percent range.

“A business can better hear its employees through AI, both for better listening and for augmenting people’s ability to accomplish tasks.”

According to Rometty, not only can IBM’s AI detect that an employee may be looking to leave a position, the ‘proactive retention’ system can generate case-specific, early-intervention suggestions for managers so they can intervene – before leaving even “enters [the employee’s] mind”.

IBM uses the system itself of course, and says that, so far, the program has saved the tech giant almost US$300 million by addressing staff attrition. (We note too that the company has also recently slashed its global HR departments by around 30 percent).

Simply put, AI-equipt managers are better at assessing employee strengths, weaknesses – and flight risk – than those using traditional methods alone.

“We found manager surveys were not accurate,” says Rometty. “Managers are subjective in ratings. We can infer and be more accurate from data.”

IBM has now patented its predictive attrition program and is offering the technology to clients. While the company is keeping hush about the inner workings of the tech, its part of a larger push by IBM to redesign human resources in general. The company has also launched My Career Advisor, an AI virtual assistant that provides employees with upskilling recommendations and training advice, and Blue Match, which assess employee skills and pairs them with matching job openings.

“AI can take all the data – structured and unstructured – and generate hypotheses, reasoned arguments and recommendations,” says IBM.

“Advances in artificial intelligence are already improving how businesses understand what their employees are telling them through their conversations, sentiments and actions… A business can better hear its employees through AI, both for better listening and for augmenting people’s ability to accomplish tasks.”

But for those still weighing up their next investment in AI, HR, or both, what’s the quick fix to better retain staff?

A recent research paper from Stanford might hold the answer. The study, More Money, More Problems: Expectations, Wage Hikes, and Worker Voice, finds that simply supplying employees with an opportunity to communicate their satisfaction levels – both positive and negative – can reduce turnover and absenteeism, even for the most embittered employees.

The researchers partnered with manufacturing firms in India. In the face of a disappointing wage increase, employees were interviewed twice, first to gauge their frustration levels at the low wage bump and secondly to record their general employment satisfaction (or lack thereof). That ‘voice’ data was then mapped onto attendance and resignation data over the following months.

The research revealed that those who took the opportunity to speak about job dissatisfaction had a 20 percent lower resignation rate than those who didn’t.

“A worker’s utility increases when she is able to communicate her dissatisfaction to her employer,” says the report.

“And the ability to lodge complaints effectively may generate positive changes in the employment relationship… Through these two channels, voice essentially functions as non-wage compensation.

“As a result, turnover should decrease when workers can – either individually or collectively – meaningfully communicate their dissatisfaction with their employer.”