Improve the Employee Experience for Customer Service Agents and Supervisors Through Conversations Informed by Predictive Analytics
Creating rewarding employee experiences and retaining employees is key to running any business. Using machine learning as a retention enabler is the focus of this blog post.
While appearing on CNBC in 2019 to announce a new tool IBM had created—with AI, machine learning, and predictive analytics—to identify employee flight risk candidates, former IBM CEO Ginni Rometty said, “The best time to get to an employee is before they go.”
In this blog post, we’ll explore how BPOs can use machine learning and predictive analytics to retain more contact center employees by assessing frontline employee attrition risk and leveraging that knowledge to create an employee experience that often addresses their unique needs.
- Employee retention in the 2020s.
- Costs of replacing contact center employees who leave.
- Modeling data to identify employee attrition risk patterns.
- Predicting and reporting employee attrition risks.
- Intervening to retain employees.
- Measuring intervention outcomes.
- Calculating the impact of intervention on P&L.
- A real-world example.
Employee Retention in the 2020s
Since the onset of the pandemic, employee retention has become an even greater challenge than it was when Ginni Rometty appeared on CNBC in 2019.
A global survey by PwC revealed that, across all lines of work, one in five employees expected to find a new opportunity in 2022.
Digitally savvy BPOs use every tool at their disposal in an attempt to retain their frontline customer experience team members.
Costs of Replacing Contact Center Employees Who Leave
Direct and indirect costs add up quickly when replacing contact center employees who leave voluntarily.
Direct costs include recruiting, onboarding, and training each new employee. It can take months to hire a new employee and train them to become fully versed in a new customer experience position.
Indirect costs become factors the moment an employee departs. That’s when the employer loses the employee’s skillset, everything the employee learned about the company and the CX program they supported, as well as internal processes when working with their team and stakeholders.
Moreover, teams have to cover the gap left by the departing employee until a new employee is onboarded and up to speed.
When a team member departs, their loss can also lower morale. When an experienced worker leaves—especially one in a supervisory role—indirect costs rise even higher.
Watch Our Webinar
iQor SVP IT Joe Przybylowski, Data Scientist Andrew Reilly, and VP Operations Terri Robertson explain how iQor uses machine learning to help reduce employee turnover.
Modeling Data to Identify Employee Attrition Risk Patterns
Organizations have long used annual surveys to measure employee sentiment. Surveys can provide a limited sense of how employees feel about working for an organization to help guide general improvements. Annual surveys are limited in their ability to identify individual employee sentiment and predict which specific employees are likely to leave, for several reasons:
- Many workers don’t complete the survey.
- Some workers might respond with what they think is the “right” answer instead of how they really feel.
- Some organizations only use anonymous surveys, so the organization doesn’t know which employees fit the profile of a flight risk.
For a more accurate snapshot of individual employee sentiment, quick surveys conducted at regular intervals that identify employees are often more effective.
Training the Machine Learning Model
Data Scientists can use machine learning to analyze harvested data from former employees’ surveys and create profiles of employee sentiment that suggest when a worker is likely to voluntarily separate from the BPO. With these profiles, data scientists can create an employee attrition risk training model for a machine learning algorithm. As new data is compiled from existing workers and as more workers depart, machine learning automatically updates the algorithm to make it more accurate.
Adding More Data Sources for Modeling Accuracy
While using a single source of data may provide some guidance, there’s always a risk that the model is biased in some way, or that the sample size is too small. Using multiple data sources makes the model more accurate.
Data scientists follow a formal process to identify and validate potential data sources based on elements of a worker’s environment, including how they interact with coworkers and supervisors.
Environmental data sources might include:
- Login and logout times.
- Frequency of breaks.
- Complexity of the tasks assigned.
- Coaching interactions.
Data scientists diligently explore other data sources through cross-functional analysis of processes.
Predicting and Reporting Employee Attrition Risks
Using multiple data sources minimizes bias built into a single data source. With enough data, predictive analytics (a subset of machine learning) can forecast which employees are the most likely flight risks and why they are at risk of leaving.
Once data analysts have an accurate view of which employees are likely to voluntarily separate—and the probable causes for their separations—they share their findings with the employees’ managers. These reports get to the heart of who’s at risk and why, and spare managers the tedious task of having to decipher the data.
The manager then determines the next step to take with each employee. Knowing the likely reasons an employee is an attrition risk gives managers a relevant starting point when they intervene proactively in their attempt to retain the employee.
Intervening to Retain Employees
Knowing why an employee is considering leaving can enable a manager to determine the best approach to resolve the employee’s concern, create a better experience for them, and help them remain on the team.
When the manager is empowered to take an empathetic approach to intervening with the employee, they help the employee understand their commitment to resolving the concern.
For example, if predictive analytics identifies an employee encountering scheduling concerns that make their job logistically challenging, the manager can work with the employee to design a more flexible schedule that better meets the employee’s needs.
If, for instance, predictive analytics identifies an employee seeking additional career growth opportunities, the manager and employee can develop a plan to support the employee’s aspirations to learn and grow within the organization.
With these and many other examples, when a manager knows the likely reasons an employee might voluntarily separate, they can tailor their approach to address the employee’s individual needs and create a better employee experience that recognizes their value to the organization and keeps them on the team.
More Real-World Examples
Learn How Machine Learning Can Help You Reduce Employee Turnover
Check out our webinar featuring iQor SVP IT Joe Przybylowski, iQor Data Scientist Andrew Reilly, and iQor VP Operations Terri Roberts to discover how to harness technology to engage employees and reduce turnover. Get the details directly from the experts responsible for this program.
Measuring Intervention Effectiveness
To determine the efficacy of interventions, companies run tests. Among a group of employee churn candidates, a portion is placed in a control group to measure the differences in outcomes between those who receive interventions and those who don’t.
Calculations over time have proven how much more effective intervening is than not intervening to retain employees predicted to be at risk of voluntarily separating through the intelligent model explained in this blog post. This can result in significant boosts in contact center employee retention as well as improved overall employee experiences, including career advancement for retained employees.
Bottom Line: Retaining Customer Experience Agents Improves CX
Experienced customer service agents build relationships with their teammates, serve as mentors to newer agents, treat their customers with the care and respect they deserve, and champion the brand they represent.
With machine learning and predictive analytics, BPOs can positively affect frontline employee retention and keep their employees and customers smiling.
Experience the Best in Data Analytics
iQor’s analytics as a service offering uses a combination of iQor’s proprietary speech analytics platform, cloud computing, machine learning, artificial intelligence, and data analysis to develop custom interventions for identified areas in need of improvement along the customer journey. The results produce targeted improvements for the employee, customer, and client.
iQor is a business process outsourcing company ideally suited to help brands create amazing customer experiences. iQor provides a comprehensive suite of full-service and self-service scalable offerings that are purpose-built to deliver enterprise-quality CX.
Our award-winning CX services include:
- A global presence with 40+ contact centers across 10 countries.
- A CX private cloud that maximizes performance and scales rapidly across multiple geographies on short notice.
- A partnership approach where we deploy agents and C-level executives to help maximize your ROI.
- The perfect blend of intelligent automation for scale and performance coupled with an irresistible culture comprised of people who love to delight your customers.
- Virtual and hybrid customer support options to connect with customers seamlessly, when and where they want.
- The ability to launch a customer support program quickly, even when you need thousands of agents ready to support your customers.
- A best-in-class workforce management team and supporting technology to create a centralized organization that can better serve your entire business.
iQor helps brands deliver the world’s most sought-after customer experiences. Interested in learning more about the iQor difference? If you’re ready to start a conversation with a customer experience expert, contact us to learn about how we can help you create more smiles.