Does better data mean better employee performance and organizational outcomes? That’s the implication of the current emphasis on big data and the use of metrics in HR, but the answer isn’t an easy “yes.”
To see what I mean, consider your local schools. When teachers are evaluated and paid on the basis of students’ test scores, performance on tests typically improves. The moral: Data works. Long live data!
But research also shows that higher test scores don’t necessarily translate to greater student mastery of the material. In other words, teaching methods that are effective in improving test scores may not be the best for increasing students’ knowledge. The moral: Data doesn’t work. Down with data!
Teaching is a great example of the strengths and shortcomings of data-based performance assessments because, in a sense, teachers are both frontline workers (when actively teaching in the classroom) and executives (when they write lesson plans and develop teaching and classroom strategies). In their role as line workers, teachers can be expected to respond to whatever metrics are applied to them. But simple metrics such as test scores may not detect the difference between teaching strategies that increase students’ knowledge and those that don’t.
A lot of us have jobs like that—some of our work leads to easily measured outcomes (sales volume), while some is much harder to quantify (solving a complex technical issue while easing customer frustration). With the rise of eHRM—electronic human-resource management—it becomes easier than ever for organizations to automate the collection and analysis of employee data. But this also means that it becomes easier to rely on data that organizations can conveniently collect and analyze. Behaviors and aspects of performance that aren’t easily quantified and captured in eHRM can become neglected.
For example, an organization that measures only the number of cases a customer-service rep handles per day may overlook the value of an employee who is capable of winning over an agitated customer. Consider also the use of workload-scheduling software for maintenance employees or physicians. These systems can increase overall operational efficiency and employee performance (measured as number of service calls completed or patients seen), but what happens if the system doesn’t account for the complexity inherent in different jobs? High-performing experts can be penalized for taking on complex assignments.
When you over-objectify or oversimplify the measurement of performance, you risk missing the richness of what makes that job special—or complex—or what makes each person’s contribution unique. Yet, for many managers, this duality is not apparent. Managerial knowledge and skill in applying metrics has not kept up with organizations’ ability to create them. Managers often don’t have the time or knowledge to understand the limitations of the metrics they apply. Instead they rely on easily obtained “objective” data from the system and ignore the less quantifiable and more complex aspects of performance.
Employees will engage in the behaviors easily captured through the system and ignore those aspects of performance that aren’t considered. That’s why organizations need to continually assess whether the data they’re collecting is truly relevant to the broader organizational objectives.
My hunch is that HR is moving toward an era of better data. What do I mean by better data? Take, for example, Sabermetrics and its use in Major League Baseball. Before Sabermetrics came along, few people imagined that the conventional thinking about baseball could be upended by arcane statistics such as wins above replacement.
Before we can develop a metric similar to wins above replacement for employees, we have to define key organizational and employee performance outcomes and determine how they relate to employee behaviors. The challenge is that we still don’t know what these metrics will look like or whether they will fully reflect performance.
Along with better data, we need to develop a more nuanced view of human qualities and human potential. Can we not only accept, but embrace, that some behaviors may not be reducible to easily quantifiable metrics, and that no amount of data can fully capture all of your, or my, best performance qualities? In a world that is increasingly driven by quantitative analyses of employees and performance, we need to find ways to efficiently incorporate both the quantitative and qualitative aspects of performance.
