The cost of somebody leaving their job and being replaced can cost a company up to 200 per cent of the post’s annual pay – not to mention the loss of firm-specific knowledge, skills and relationships.

And that is if you can replace them. More than half of Fortune 1000 firms report that unfilled STEM jobs (science, technology, engineering and mathematics) have led to lower productivity for the firm. While in China and Mexico the turnover rates of 60 and 100 per cent among its manufacturing industry threatens global supply chains with not only higher labour costs, but an increase in product defects and production shortfalls.

So, the simple answer is to stop talent leaving. The slightly trickier task is creating an environment where people don’t want to leave.

Before companies can create such an environment they need to understand just why people leave their jobs – and research has shown it is not all about money.

1 Low job satisfaction

Poor attitudes about one’s job remain an important consideration. Leaders won't be surprised to learn that pay does matter when it comes to job satisfaction, but not nearly as much as workers being treated fairly and valued for their efforts.

Pay level and satisfaction with their salary are actually relatively weak indicators of a worker’s decision to leave. Much more important is job satisfaction overall, which takes into account the whole working environment.

Employee engagement is something that all companies are striving for and analysing as a complement to job satisfaction. A key component of engagement is that workers feel the organisation cares about them and values their contributions. Managers need to work hard at making sure staff are appreciated, but this needs to be done in a genuine way.

Giving workers autonomy, variety and challenging goals with the appropriate support have all been shown to foster greater job satisfaction and engagement. Plus, showing workers how their job contributes to the greater mission of the organisation also creates more commitment and satisfaction among staff.

If companies do not give workers freedom and challenging work, with clear goals and fair rewards, they will become dissatisfied and look for employment elsewhere.

2 Shocks

Traditionally it has been thought that staff become slowly tired of their work and the routineness of it, then start to look elsewhere for a new and more challenging job. But more often than not, research has found, that people leave their job because of the shock of a specific event, which forces them to re-evaluate their position.

This shock can come in the form of not gaining a promotion that they thought they were going to win; a more negative performance appraisal than expected; or a smaller bonus than they had anticipated.

A shock does not have to be job-related either. Workers can win the lottery or suddenly come into a lot of money, which would cause them to think 'do I really want to work here?' They might gain a new qualification or credential or they or their partner could become pregnant.

Research has shown these type of events are really common predictors of why people leave their jobs. A manager focused on their employees, can figure out which types of shock are likely to happen and can can come up with scripts to plan ahead for that event.

For example, firms can plan whether to make a counter-offer with a wage rise or look into its policies if, for example, pregnancy is seeing a lot of talented people leave.

3 Embeddedness

Another perspective focuses more on why people stay than why they leave. Staff who are embedded in relationships and networks inside and outside the organisation are more likely to stay.

People over time become enmeshed in a web of contacts and communities in the company through friends, work teams, committees, social events and lunchtime activities.

For some they build-up a community of friends that it might be hard to give up. This is not just the workplace but embeddedness in the wider community, town or city. Making friends with the neighbours, buying a house, doing voluntary work or joining sport teams makes people more embedded.

Even if their supervisor is abusive and they are losing sleep plus having health problems, there is evidence people still won’t leave if they are so embedded. Thus, on the flipside, those who are less embedded are more likely to leave.

4 Turnover contagion

There is growing evidence that turnover is contagious in the workplace, just like emotions and job attitudes are contagious. If a group of workers become dissatisfied and start job searching then other people in the organisation are more likely to search for another job.

Companies can now measure communication networks in an organisation, they can see who talks to whom, so if somebody decides to leave they can see who they are connected to and intervene.

It might be that some people are very central to a network, as can be seen by the pattern of emails they send to and receive from the group, so if they leave that would present a real problem to a company and the real possibility of it spreading with many people following them out of the organisation.

In order to do anything about these four reasons for employee turnover it is vital companies collect data. Most companies, however, rely on exit interviews, which only give information after people have decided to leave, or annual employee surveys, that can be challenging to act on.

Thus, more and more organisations are collecting much broader types of data to create algorithms that will predict who is going to leave based on job attitudes, shocks, embeddedness and their network.

These can be intrusive, such as tracking and mapping email traffic. Firms are usually not looking at the content, they are just tracking where the emails go and their frequency.

Employee ID cards are also a source of data, allowing companies to track workers’ movements as they swipe it to get into rooms, through entrances and in lifts. Workers may have a pattern of movement and if that suddenly changes - such as they start leaving the building at 2pm every day - it will flag an alert to the company for it to intervene.

All this data can be used as a dashboard for managers on each of their direct reports, so that when somebody goes ‘red’ they can respond in some way. Some firms are also using social media data, with updating of a worker’s LinkedIn profile an obvious red alert.

There is simpler data managers can use as research has found people are more likely to make a job change around an anniversary, such as five years into the job, or even on birthdays.

Some amount of employee turnover is going to be expected as companies look to move out the ‘energy sappers’ and low performers, but being aware of these four major reasons for staff leaving will help managers intervene and keep hold of the talent their firm needs to succeed.

Further reading:

Hom, P. W., Allen, D. G. and Griffeth, R. W. (2019) "Employee retention and turnover: why employees stay or leave", Routledge, .9781138503816.

Holtom, B., and Allen, D. G. (2019) "Better ways to predict who's going to quit", Harvard Business Review.

David Allen is WBS Distinguished Research Environment Professor at Warwick Business School and Associate Dean and Professor of Management and Leadership at TCU Neeley School of Business.

For more articles on Leadership sign up to Core Insights here.