Calling it quits! Tech start-up invents algorithm that predicts when employees will leave their jobs

Calling it quits! Tech start-up invents algorithm that predicts when employees will leave their jobs

  • Tech start up Zegami, founded at Oxford University, helps to spot unhappy staff
  • Co-founders Roger Noble  and Sam Conway believe algorithm is first of its kind
  • Users feed the programme with employee characteristics such as age and salary 
  • Zegami estimates average cost for losing an employee is more than £105,000

A tech start up has invented an algorithm to predict when employees will hand in their notice. 

Researchers a data visualisation company Zegami, founded at Oxford University, created the new software to help bosses spot unhappy staff.

Users feed the programme with various characteristics of employees, including their age, salary, job history, education and output.

Co-founders of Zegami, Roger Noble (left) and Sam Conway (right), have invented an algorithm to predict when employees will quit their jobs 

The algorithm then highlights staff members who are ‘at risk’ and gives a time period for when they might leave.

This is based on research from Oxford University, which shows what keeps employees happy in jobs.

Zegami, whose clients include Major League Baseball side the Pittsburgh Pirates, then presents the data in easy to understand visuals.

Co-founder and chief executive Sam Conway believes this is algorithm is the first of its kind.

Co-founder and chief executive Sam Conway said by looking at the data you can understand the personality of your employees and can persuade them to stay by raising salary or allowing them more time with their family

Co-founder and chief executive Sam Conway said by looking at the data you can understand the personality of your employees and can persuade them to stay by raising salary or allowing them more time with their family 

‘By seeing the data, you can understand the personality of your employees,’ he said.

‘This is incredibly valuable and adds an extra dimension to your decision.

He added: ‘You can group together all of the people who are ‘at risk’ and the cost to my business is this.’ 

The data also links to other employees so bosses can spot if there is a bad apple in one team.

Mr Conway said it ‘seems pretty obvious’ to analyse the date if there are six people in the department and four want to leave however many companies don’t take this initiative. 

Roger Novel using Zegami which can highlight staff members who are 'at risk' and allow them  a time period for when they might leave

Roger Novel using Zegami which can highlight staff members who are ‘at risk’ and allow them  a time period for when they might leave

A business can decide whether to raise a staff member’s salary or give them more time with their family to stop them handing in their notice. 

‘We can do things without the person even knowing, which makes them want to stay,’ Mr Conway explained.

Zegami estimates that the average cost for a business losing an employee is more than £105,000.

This is based on the average salary for someone in full-time work, hiring costs and productivity losses over several months while the replacement becomes as effective as the original employee.

‘Our aim is to make this data accessible for everyone,’ Mr Conway added.

The tech business is looking to branch into artifical intelligence as its next step.  

‘What we want to do is to harness this power to improve the search experience and democratise data analysis,’ Mr Conway added. 

How bosses know if employee is likely to leave 

To find out the likelihood of an employee handing in their notice, the programme takes in a range of considerations to calculate a risk factor. 

  • Country employee is from  
  • Age
  • Team they most often work in 
  • Interpersonal relationships 
  • Salary rate 
  • How long they have been at the company 
  • Satisfaction 
  • Behaviour  

Co-founder of Zegami Sam Conway said: ‘Change or disruption in the workplace can have a knock on effect on an employee as well as interpersonal relationships with colleagues. 

‘When you have the data you can see who works together, in what team and how long within that department people have worked.

‘By seeing how long people have been in the company you can identify a pattern in the data where potentially all the new employees are at risk of leaving or all the old employees.’  

Read more at DailyMail.co.uk