.jpg)
Every employer will have experienced being blindsided by a resignation they never saw coming. And recent research by PwC shows that a quarter (26 per cent) of employees are looking to move jobs in the next 12 months, up from 19 per cent last year. All too often, high-performing employees make the decision to leave without much warning, leaving companies feeling that they could have done something to prevent it had they spotted the warning signs.
Determined to avoid recurrences in the future, organisations will try to increase monitoring of the various factors liable to push employees to leave. In the age of digitisation and AI, it can be tempting to use the efficiencies of HR technology to forecast flight risk among the workforce, but this is not without its dangers. HR systems – including and especially the more comprehensive ones with sophisticated machine learning algorithms – generally do not have a high degree of accuracy when it comes to making such predictions. The data that such software holds is relatively sparse and overlooks the myriad complexities that factor into an employee’s decision to walk away from a role, some of which are too nuanced for a digitised system to register and decipher. Other, more personal, considerations are not even available to the employee’s manager, let alone to HR technology.
As such, it is unwise for employers to rely solely on HR software to tackle flight risks within their ranks. Instead, they should look to deploy other methods, with more human elements, and only use digital tools to supplement this. So, how can businesses balance the human approach to personnel management while bolstering the capabilities of HR technology?
Soft skills or software – or both?
When it comes to pinpointing and preventing flight risk, the human touch is vital. Even the most bespoke, cutting-edge HR technology cannot replace managers having regular, open conversations with their charges. These can have an underestimated impact on improving the employee’s feelings towards work, and the company’s chances of spotting any problems early on.
Additionally, a machine learning algorithm does not possess the soft skills that are vital in dealing with potential flight risk, such as caution, common sense and emotional intelligence. An HR system is also not privy to the day-to-day niceties and interactions in an employee’s work life, any external personal factors or the degree of loyalty they feel towards the company. However, it does provide an additional resource to record other useful digital factors, such as employee morale and salary benchmarks.
Besides the lack of accuracy resulting from not having a complete picture, surfacing a flight risk prediction from HR tech can be risky. There is potential for a manager to alter their behaviour towards a ‘high risk’ employee unnecessarily, including passing them up for a training course or promotion. The bias can also operate in the reverse direction, with managers giving these employees preferential treatment in a bid to make them want to stay. By predicting via software, companies have no way of ever verifying their predictions, even if an intervention induces an employee to stay. They are better off adopting an alternative, more personalised approach to this issue, rather than chasing the accuracy and precision that only a human touch can deliver.
Tracking over predicting
Domain expert-led indicators can be a valuable technique to regularly track common factors impacting flight risk, enabling employers to spot and gain insight into an employee who may need some attention. They also help avoid singling out any individuals or raising alarm bells. These could operate at two levels – personal and general. Personal indicators would regularly track factors such as the number of disruptions experienced by an employee, whether that be manager or team changes, workload volumes, sickness frequency or holiday usage. At the same time, general indicators would aggregate employees who have similar characteristics (departments, job levels, skills or tenure) into ‘cohorts’ and track patterns at a macro level, such as the average salary and longevity of employees in different roles, including comparisons to industry benchmarks.
Of the social aspects affecting an employee’s choice to leave or stay, their relationship with both their colleagues and their manager is often a key contributing factor. When good colleagues depart from a company, an employee’s social links diminish, as does their sense of stability and commitment towards the business. This is something that would be missed by even the most sophisticated HR systems, but which the ‘disruption’ indicator would regularly and consistently track across all employees. Having said that, digital tools can be a useful, secondary add-on for holding and analysing this data once it’s been gathered in this way. The information should then influence decisions around manager interventions to enhance an employee’s experience, while also being used to improve long-term retention strategies.
If gleaned and handled in the right way, ‘flight risk’ insights can be an invaluable resource for organisations looking to create a better workplace experience for all employees. When dealing with such a sensitive issue, the human component of HR is key and employers should not underestimate the value of frequent, regular and open dialogue. Any essential information and changes can be fed into ongoing indicators providing oversight of all the common factors that impact flight risk and this can then be bolstered through tech.
Jeanette Wheeler is chief HR officer at MHR