How to Correlate Disparate HR Data for Improved Talent Decisions

​The average HR department is awash in unprecedented amounts of data generated by its core HR information and talent management systems. But despite that wealth of data, many departments remain knowledge-poor due to a fundamental shortcoming: an inability to combine and correlate different HR datasets in ways that create improved workforce insights.

Aligning the right data from a performance management system and an applicant tracking system (ATS) can generate quality-of-hire metrics so that recruiters can recalibrate sourcing strategies to target candidates more likely to succeed on the job. Connecting answers on employee engagement surveys to post-survey behaviors can produce insights like whether those who said they were going to leave a company actually left.

Such analysis requires that HR pull data out of silos scattered around the organization and centralize it for side-by-side comparisons. Organizations—and industry vendors—are taking steps to make that possible by investing in the infrastructure, data warehouses and analytics tools.

"When I talk to customers, they say one of their biggest challenges is the time it takes to consolidate data from an ATS, HRIS [HR information systems], or performance or engagement system onto one platform," said Rajeev Behera, CEO of Reflektive, a San Francisco-based tech company. "If you add performance metrics to almost any other HR dataset, you usually create valuable insights, but that type of correlation doesn't happen often enough in organizations," typically due to infrastructure constraints or a lack of data science skills in the HR function.

Interest in Advanced Analytics

Recent research suggests HR professionals have a growing thirst for such analytics initiatives. Deloitte's 2018 Global Human Capital Trends survey results found that 84 percent of respondents viewed people analytics as important or very important, making it the second-highest-ranked trend in the survey. More than 70 percent of respondents also reported being in the midst of major projects to analyze and integrate workforce data into their decision-making.

Some industry vendors are acting on that trend by enabling new ways of correlating data on their platforms. Kronos, a Lowell, Mass.-based provider of workforce management and human capital management (HCM) technologies, lets users combine data from its time and attendance, employee scheduling, absence management, and other modules to make more-informed workforce decisions.

The platform can calculate a "reliability" score for employees that managers can use when deciding how to schedule work shifts. Reliability scores are devised primarily on how many times employees clocked in late and how many absences they had in the past; the score might also factor in performance reviews and whether employees have picked up shifts from co-workers in the past.

The manager of a convenience store, for example, might only have two to three people at a time working a given shift, making it essential that workers show up on time. "That manager could review the reliability scores of workers available for the shift and predict whether they might have problems with people calling in sick or showing up late," said Bob DelPonte, vice president of the HCM practice group at Kronos.

Reliability data also can be correlated with recruiting data when making hiring decisions. "When store managers are hiring, they might see that current employees who live within five miles of the store had higher reliability scores than those who lived 15 or 20 miles away," DelPonte said. "They could factor that data into their recruiting decisions."

Deb Wolfsen, a human resources manager with 200-employee Engineered Protection Systems in Grand Rapids, Mich., uses data correlations to track and analyze the company's recruiting initiatives and efforts to reduce turnover. Wolfsen also correlates performance data with recruiting data to analyze quality-of-hire metrics.

"We look for trends in the backgrounds of employees we've hired who have performed well," she said. "For example, we've discovered people who've formerly worked at the front desks of hotels have done well in our customer-facing roles. So we do searches for applicants with similar characteristics."

Some HR leaders also are now combining datasets to create "quality of life" scores for employees. Someone who hasn't taken vacation time recently or who has been working a lot of overtime could be at risk of burnout or of leaving the organization. How might that data be used in workforce decisions? When determining who should get time off when there is only one slot open on a work schedule, a manager could look for the employee with the lowest engagement score and give them the day off, which might help improve job satisfaction.

Maximizing Engagement Survey Data

Experts say integrating systems to combine HR data is one thing, but extracting valuable insights from those initiatives is another. The latter typically requires a certain level of data science competency. "It's easy enough to connect an ATS to your HRIS, but what do you do with the data at that point?" said Chris Butler, CEO of people analytics provider One Model, who formerly worked in HR analytics with SAP SuccessFactors. "You need to look beyond the core set of data to provide additional context to help managers understand what's going on with your workforce."

Butler believes engagement surveys represent one of the biggest missed opportunities in correlating HR datasets. Data from those surveys can be connected to almost every other data point HR has about employees, he said. Some companies are hesitant to do this because they don't know how to leverage confidential survey data correctly and maintain the privacy provisions they promise to employees in data collection.

Vendors that administer employee engagement surveys often collect some form of employee identifier, Butler said, and with the right controls and privacy protections, organizations can confidentially link survey results to actual employee outcomes.

Butler said connecting engagement survey responses and post-survey behaviors can generate insights like:

• Did people who responded that they lack the opportunity for training actually take a training course once it was offered?
• How do employee rewards affect subsequent engagement scores?
• Did people who expressed a lack of advancement opportunities on surveys leave the company for a promotion?

Dave Zielinski is a freelance business writer and editor in Minneapolis.