Employee retention and attrition

How data can help increase retention and reduce staff turnover

Data driven employee retention

With the war on talent raging as never before employee retention and attrition are paramount themes for almost every organization.

As a result many (HR-)organizations have made it their priority to improve employee retention and reduce employee attrition. Equally, many (HR-)organizations don’t achieve their retention and attrition objectives.

Therefore it is time for a new approach to employee retention and attrition; a data driven approach. This page lays out:

Why a data driven approach to employee retention and attrition is superior to a conventional one
How a data driven approach to employee retention and attrition should be implemented
What organizations can think of when they want to pivot from conventional retention practices to data driven retention
Case examples of successful data driven employee retention and attrition initiatives
Answers many Frequently Asked Questions (FAQs) about retention and attrition in relation to data

More information? Feel free to contact us!

Why is data driven employee retention and attrition important

Data driven HR is obviously about data. While data takes a prominent position in data driven HR another key element of data driven HR (also known as HR Analytics or People Analytics) is structured problem-solving.

In practice, many organizations still rely on assumptions, gut feelings and don’t go beyond looking at averages when assessing a challenge. A data driven approach to Retention and Attrition really helps to narrow down on those employee segments that are at risk and require attention. It helps organizations to pivot from a generic blanket approach to a laser focused approach which can vary from one employee segment to another, thereby making a bigger impact with equal or fewer costs.

How to approach data driven employee retention and attrition

Sounds interesting? Implementing a data-driven approach to employee retention and attrition can be done through the following steps:

1. Collect data:
The first step is to collect and store data on employee retention and attrition. This can include data from operational HR systems on employee demographics, hiring, promotions, transfers and separations. Ideally, the data set should also include employee survey data. For instance from the engagement, onboarding and exit survey.

2. Visualize the data:
Once the data is collected, it should be visualized in an Employee Retention and Attrition dashboard. A dashboard will allow HR to establish in which employee segments retention and attrition is truly a problem. Kindly find an example HR Retention and Attrition dashboard below.

Need help with building an Employee Retention / Attrition Dashboard? Contact us!

3. Define required insights:
Once the employee segments with a retention/attrition challenge have been identified, analysis questions can be defined for each segment. Or in other words, which insights are required to improve retention and reduce attrition? For example:

Within the groups with increased attrition, what factors are related to retention/attrition?
Why are people within these groups leaving the organization?
Where are these people leaving to?
What is the flight risk of people within these groups?
Which teams can expect unnatural attrition in the future?
What are potential mitigating actions for the people who are at risk?

4. Apply Data Science:
With employee segments and analysis questions defined Data Scientists can start cracking. Data Scientists will use Machine Learning techniques to answer the abovementioned questions, or they will go one step further and can build algorithms that support HR in its daily operation. Using predictive or even prescriptive data science techniques, Data Scientists can build an algorithm which indicates in which teams the chance of attrition is the highest. The algorithm can even provide suggestions to HR about potential interventions to prevent attrition in the team.

5. Develop & Implement retention strategies:
Using the insights gained from the data analysis, organizations should develop retention strategies that address the identified risk factors. These strategies should be tailored to the specific needs and preferences of the different employee groups. Once the retention strategies are developed, they should be implemented across the organization. Change and communication management is important in this phase. Share the findings, insights and plans with all the stakeholders involved, including employees and management, to ensure buy-in, transparency, and accountability.

6. Monitor and evaluate:
The effectiveness of the retention strategies should be monitored and evaluated over time. Are we hitting the mark, or do we need to calibrate our approach? This is where the Employee Retention/Attrition dashboard comes into play again.

What are examples of data driven employee retention and attrition interventions?

As the reasons for employee attrition can differ per sector and even per company, but also because organizations can be in a different HR Analytics maturity stage, the solutions (HR-)organizations implement around data driven employee retention and attrition also vary. Below are some common data driven employee retention and attrition practices organized in three categories: surveys, analytics/data science and dashboarding/metrics.

Surveys

Onboarding survey: A well-designed and effective onboarding journey can help to set new employees up for success in their new role and can increase their engagement and commitment to the organization. This in turn can lead to higher levels of employee retention. Employee Journey Mapping in conjunction with an Onboarding survey delivers the insights to improve the Onboarding journey and thereby indirectly influence employee retention and employee attrition.

Engagement survey: once employees have been successfully onboarded it is important to keep track of their engagement and retention intention as these are critical leading indicators for retention/attrition. An engagement survey can be used for this purpose. The data collected with an engagement survey can be used to identify drivers of employee engagement, which can help organizations to develop targeted retention strategies.

Equally it can be used to develop predictive algorithms, e.g. an employee flight risk prediction model which HR can use to have a targeted and meaningful dialogue about attrition prevention.

Exit survey: despite all the efforts to prevent attrition it is hard/impossible to prevent all employee turnover. Therefore it is important to also run a continuous exit survey. The survey is designed to gather information about the employee’s reasons for leaving, their perceptions of the organization, and their experience while working at the organization, but also information about where or what they are leaving for. By gaining insights into these issues and trends and addressing them, organizations can improve the overall employee experience and reduce the likelihood of employee turnover.

Need help with building an Employee Retention / Attrition survey? Contact us!

Analytics / Data Science

Diagnostic modelling: Data Scientists can use data to create diagnostic models that identify the factors that are drivers for employee turnover. These models can be used to formulate actions plans and policies.

Predictive modelling: Data Scientists can use data to create predictive models that can identify the factors that are most likely to lead to employee turnover. These models can be used to identify employees who are at risk of leaving and to take steps to retain them.

Cluster analysis: Data Scientists can use cluster analysis to group employees together based on similar characteristics, such as job role, tenure, or engagement level. This can help organizations to identify patterns in employee turnover and to develop targeted retention strategies for different groups of employees.

Text analysis: Data Scientists can use text analysis techniques to analyze text originating from employee surveys, performance reviews, to identify common themes or patterns in employee feedback. This can help organizations to identify areas where improvements are needed and to take steps to address any issues that may be contributing to employee turnover.

Network analysis: Data Scientists can use network analysis to identify key influencers or connectors within an organization. This can help organizations to identify employees who are vital to the organization and to take steps to retain them and/or use them as ambassadors in change programs, thereby driving buy-in amongst employees.

Deep learning: Data Scientists can use deep learning algorithms to analyze (anonymized) unstructured data such as employee emails, chats and other digital interactions to identify patterns in employee behavior that might indicate that an employee is considering leaving the company.

Need help with Employee Retention / Attrition data science? Contact us!

Dashboards/Metrics

Tracking key employee retention and attrition metrics (see box below) can help organizations to identify areas of heightened employee attrition and measure the effectiveness of retention efforts.

Ideally these metrics are included in a visual and interactive HR Dashboard that allows HR professionals to swiftly identify if and where challenges exist (scroll upwards for an example dashboard).

Example employee retention & attrition metrics

Onboarding Satisfaction
Engagement Level
Retention Intention
Involuntary Attrition Rate
Reasons for Involuntary Attrition

NB: HR should be able to view the metric values in different
dimension, e.g. Business Unit, Business Line, Team; Region,
Country, Site; Age categories; Tenure categories; etc.

Need help with defining Metrics or building an Employee Retention/Attrition dashboard? Contact us!

Case examples of data driven employee retention and attrition

General Electric (GE) uses advanced analytics to identify employees who are at risk of leaving and to develop targeted retention strategies. They use data from employee surveys, performance evaluations, and other sources to identify key drivers of employee engagement and satisfaction. They also use machine learning algorithms to predict the attrition risk and improve their employee retention strategies.

Amazon uses data-driven insights to improve employee retention. They conduct regular surveys and use data from performance evaluations and other sources to identify key drivers of employee engagement and satisfaction. They also use machine learning algorithms to predict which employees are at risk of leaving and to develop targeted retention strategies.

Procter & Gamble used data analytics to track employee engagement and identify patterns in turnover. They found that employees who felt a sense of “purpose” in their work were more likely to stay with the company. Based on this insight, they made changes to their management practices to give employees more opportunities to find meaning in their work.

Walmart used data analytics to track employee engagement and identify patterns in turnover. They found that employees who felt a sense of “community” in the workplace were more likely to stay with the company. Based on this insight, they made changes to their management practices to promote a sense of community and belonging among employees.

Jumbo Supermarkets: has deployed onboarding, engagement and exit surveys and performs data science to discover why people stay and leave and where they leave to. Subsequently Jumbo uses the insights generated with these techniques to take data informed actions and implement policies aimed at reducing attrition and improving retention. Impact is tracked with various HR Dashboards.

Frequently Asked Questions in relation to Employee Retention and Attrition

How to calculate employee retention?

Retention rate = (Number of employees at the end of the period – Number of new hires during the period) / Number of employees at the beginning of the period

For example, if a company had 100 employees at the beginning of the year, hired 10 new employees during the year, and had 92 employees at the end of the year, their retention rate would be: (92 – 10) / 100 = 0.82 or 82%

This means that 82% of the employees who were employed at the beginning of the year were still employed at the end of the year.

This formula can be used to calculate retention rate over any period of time, such as monthly, quarterly, or annually.

It’s worth noting that measuring retention rate helps you understand the percentage of employees who stayed with the organization over a period of time. It’s a good way to understand how much employee turnover your organization is experiencing, but it doesn’t necessarily tell you the reasons behind why employees are leaving, or how to improve employee retention. To understand these factors, onboarding, engagement and exit interviews can be used in combination with data science techniques.

How to calculate employee attrition?

Attrition rate = (Number of employees who left during a period / Average number of employees during that period) x 100

For example, if a company had 100 employees at the beginning of the year, and 15 employees left during the year, the attrition rate would be: (15 / 100) x 100 = 15%

This means that 15% of the employees who were employed at the beginning of the year left the company during that year.

This formula can be used to calculate attrition rate over any period of time, such as monthly, quarterly, or annually.

It’s worth noting that measuring attrition rate helps you understand the percentage of employees who left the organization over a period of time. It’s a good way to understand how much employee turnover your organization is experiencing, but it doesn’t necessarily tell you the reasons behind why employees are leaving, or how to reduce employee attrition. To understand these factors, you should use other metrics such as employee engagement and satisfaction, exit interviews, and employee turnover costs.

Is it allowed to predict employee flight risk at an individual level?

It is possible to predict employee flight risk at an individual level. Flight risk can be predicted by analyzing data such as employee engagement, job satisfaction, turnover history, and other factors that indicate an employee’s likelihood of leaving. But, it’s important to be aware of the legal and ethical considerations when doing so.

From a legal perspective, employers must comply with laws and regulations related to discrimination, data privacy (GDPR), and equal opportunity. From an ethical perspective, employers must consider the potential negative impact on employees’ privacy and well-being. Employers must consider the potential impact on employee trust and engagement and ensure that they are transparent about the process and their intentions.

If predicting flight risk at an individual level is not preferred, predicting flight risk at a higher level (e.g. team), can be considered.