People analytics won’t help much without the right metrics and a keen understanding of how they relate to employee engagement, customer satisfaction and overall performance.
HR departments are awash in new data sources, tools and survey techniques to improve enterprise performance, employee engagement and productivity. But interpreting this data in a meaningful way and creating feedback loops that produce results are not easy tasks.
Experts interviewed at the Work Rebooted conference last month in San Francisco explored some of the best practices for using HR data analytics to improve employee engagement and productivity. Some of their recommendations included customizing employee engagement surveys and psychometrics for specific teams and roles, aligning new practices and initiatives with enterprise results and gaining clarity on what engagement means in practice.
“Most surveys are poorly designed for us,” said Chase Rowbotham, head of people analytics at Genentech, a biotechnology company. “We don’t have a one-size-fits-all solution for the entire company. Scientists tend to be highly skeptical, and salespeople tend to be happy in general. We would never do a monthly pulse measurement of scientists because they would throw up their arms in protest.”
Rowbotham is also concerned about the control that various engagement and cultural survey tools maintain over the data. Survey vendors claim that this is to protect employee anonymity, leading to more accurate results. But Rowbotham feels that this can make it harder to interpret results and determine if the tool is a good fit for Genentech.
Focus on outcomes and behavior
Another HR data analytics approach looks at the activity of individuals across the company. It can include the number of calls answered in a telemarketing center, lines of code written or individuals who went through a new training program.
These are all easy to measure, but of limited value. “Businesses don’t tend to measure the hard things about what is required to achieve enterprise outcomes,” said Barry O’Reilly, founder and CEO of ExecCamp, a training provider and consultancy.
For example, if a company is trying to create a better culture, it is more useful to measure how culture relates to enterprise goals. This could include metrics like the pace of software delivery, the number of people who want to join the company or customer satisfaction.
“High-performance organizations identify what they are trying to achieve and then run experiments like sending people on training and then measuring performance improvement,” O’Reilly said.
When companies focus on the wrong measures, they can increase costs for the enterprise. For example, efforts to reduce call center duration can lead to failure demand that shows up as more follow-up calls or decreased customer satisfaction. It’s possible that the enterprise ends up paying twice to solve a customer problem. “Rather than outcome metrics, a better goal is to fix the problem every time a customer calls,” O’Reilly said.
It’s also important when using HR data analytics to focus on leading metrics rather than lagging metrics to get faster feedback. For example, employee churn is a lagging metric of engagement. A leading metric would involve quantifying some behavior associated with engagement, like hosting internal learning sessions, assisting coworkers or speaking at conferences.
Use AI to provide better feedback
As more work moves online or is quantified using the internet of things (IoT), organizations have more tools to interpret different aspects of workplace performance. “We can also analyze it using AI or other tools to extract meaning from it,” said Terri Griffith, associate dean at Santa Clara University.
At the same time, it is important to focus on using these tools to give employees meaningful feedback to build trust. “The first time you bust someone for something you see, you lose trust,” Griffith said. As a result, people end up trying to game the system rather than just doing work.
If the HR data analytics shows some employees are more productive than others, make it easier for employees to understand how their peers are getting better results. Another practice might be to alert employees when a coworker might be a better fit for a task. “The more transparent we are about work and outcomes, the more likely we are to trigger people to have those conversations,” Griffith said.
Assess the true value of engagement
Another good practice is to start with the value that employees generate rather than the front-line cost when making a case for a workplace change to finance or the C-suite, said Kate Lister, president of Global Workplace Analytics, an HR consultancy. “If you take revenues of a financial institution by salary of employees, the true value ends up being an average of about six to one.” Admins may generate a value about equal to their salary, while other people will generate a value of 20 to 30 times their salary, she said.
When applying HR data analytics to assess efforts to improve employee engagement, it’s easier to focus on absenteeism but harder to measure presenteeism, which includes the time lost by workers who are not performing at capacity because they are cold, angry or dissatisfied, according to Lister.
But new IoT devices for measuring employee behavior and physiology, like stress levels, interaction patterns or behavior, could help measure presenteeism, she said. For example, Hitachi is already doing preliminary research on new sensors for its workers, and other companies are experimenting with “sociometric” badges. “The environment will sense our state and be used to increase understanding and make better decisions,” Lister said.
Clarify the metrics from HR data analytics
Enterprises might be able to get better results with survey techniques and data analysis by using prior research as a starting point, said Shreya Sarkar-Barney, CEO and founder of Human Capital Growth, an HR consultancy. It’s easy for HR departments to conflate concepts like engagement and satisfaction or to get entranced by marketing claims for new measurement techniques.
Medical science is built on a pillar of evidence-based research that considers individual studies, as well as meta-analysis of studies from multiple perspectives. In the business realm, enterprises tend to get caught up in individual success stories or vendor research about culture and engagement, according to Sarkar-Barney.
Instead, an evidence-based approach might look across individual studies to build a more solid foundation for business decisions based on HR data analytics.
One challenge is that the psychological sciences and employee engagement proponents seem to have different definitions of engagement. In psychology, engagement is a measure of absorption, dedication and commitment, according to Sarkar-Barney. “The tools used most often in business for measuring engagement are based on the Gallup survey that measures discretionary behavior and going above and beyond your duty to support your work,” she said.
Survey tools that have shown promise in measuring engagement include the Utrecht Work Engagement Scale and the Job Engagement Scale. In contrast, the Gallup approach is much better at predicting satisfaction and employee churn, Sarkar-Barney said.