Taking the Long View in a Big-Data World
From an HR perspective, global organizations are eager to tap their stores of Big Data to unlock workforce insights, but are they more likely to get lost in the fog of Big Data analysis?
By Brian J. Kelly and Julia Howes
If 2013 has produced a breakthrough technology phrase, it is Big Data, a fairly vague but forceful term that shows up predictably in just about every article, column or blog touching on the subject of capturing and analyzing the ever-rising tide of digital information that flows around the world.
Indeed, as Steve Lohr wrote in the New York Times recently, Big Data is a catchall that means three things: "First, it is a bundle of technologies. Second, it is a potential revolution in measurement. And third, it is a point of view, or philosophy, about how decisions will be -- and perhaps should be -- made in the future." From an HR perspective, global organizations are eager to tap their stores of Big Data to unlock workforce insights, but are they more likely to get lost in the fog of Big Data analysis?
That's an appropriately big question, and it calls for a refinement of the conversation, beyond the emphasis on sheer data mass and toward a more helpful view of data over time, or Long Data. Coined by mathematician and network scientist Samuel Arbesman, Long Data describes "data sets that have massive historical sweep." This makes particular sense for employers in their quest for insight into their most valuable asset: talent.
In terms of workforce analytics, then, Big Data's depth is typically focused on a point in time, while Long Data derives its strength by unpacking stories that take place over the long-term. This allows organizations to identify, segment and analyze data as part of an ongoing process, avoiding the risks associated with the perpetual habit of making purely reactive or point-in-time decisions.
Companies can benefit from a Long Data approach that includes, for example, the linking of performance management data with historical compensation practices. In this context, consider that most annual employee performance ratings aggregated across a large population match a classic bell curve, which is not, of itself, very revealing. But with the proper segmenting of employee populations and their performance over time, trends and barriers can reveal themselves more meaningfully.
This is the Long Data advantage all executives, from line-of-business to HR and C-suite leadership, should expect from the workforce-analytic technologies they invest in. Ideally, these technologies can view performance management and compensation data longitudinally in order to identify and drive a culture of sustained high performance.
Long Data's Potential
The Long Data view has the potential to fundamentally change the traditional annual review process, highlighting key areas of focus and catalyzing targeted, root-cause analysis. It should help answer strategically critical questions about how to reward high-performing employees; how to make better decisions about performance-based salary, short-, and long-term incentives; and how to better manage low performers. More broadly, it helps in the analysis of current (as well as projecting future) workforce risks and costs..
If we think about workforce issues and the employee experience within an organization, it becomes easier to identify a litany of applications for Long Data techniques. An employee's career occurs through time, after all, and need not be viewed as a series of point-in-time promotion steps. HR needs to have better insight into these through-time trends in order to understand what experiences better create and accelerate the development and competencies of employees. Long Data techniques allow us to:
* Analyze recent promotions to leadership positions, by grade at hire, to understand what balance of build versus buy is most effective for creating future leaders.
* Analyze promotions at any tenure, by pre-employment assessment and testing scores, and recruiting source, to understand not only the impact these sources have on short term quality of hire, but also longer term on the promotion opportunity of employees.
* Analyze high performing senior employees by the number and type of career steps by tenure, to understand what experiences in the organization accelerate the development of leaders.
* Analyze the implications of workforce initiatives, such as wellness programs and diversity programs, where the impact is not immediate, and can often occur in different reporting periods and by different employee segments.
Organizations understand that they can better manage and reward their employees if they apply data-driven decisions to the process, and they are ready to make investments in metrics and analytics that mine HR gold from Big and Long Data. But not all employers know the difference between workforce analytics and basic reporting, while many organizations lack confidence in the quantity and quality of their data and systems.
This may prevent them from moving forward with an analytics initiative that can benefit from Long Data. Yet only by moving forward can HR begin to prioritize the data clean-up efforts required to deliver sustained performance-management value. Companies can begin with the most critical workforce issue that needs to be addressed, and consider what data (often from multiple sources) is required to do so.
Focusing on those few data points can be the most immediately effective and valuable way to make use of workforce analytics -- a process that can proceed in phases, as the insights revealed by Long Data compound themselves. Understandably, organizations want to know if there are certain "magic" metrics, or a cheat sheet of metrics, that can accelerate their analysis and drive business results more quickly. And while there are no such short cuts, there are key areas that HR leadership can focus on to yield greater results from the data.
For example, the concept of Internal Labor Markets (ILM) is a powerful analytic tool, focusing on how organizations manage their existing labor pools to help achieve business goals. Significant insights can be derived from data that tracks how talent enters an organization, moves through it (via promotions, demotions, lateral movements, international assignments), and ultimately exits. In identifying turnover trends and/or hot spots, ILM further helps in focusing on workforce risk. But each company's ILM mapping tends to be different, and only an in-depth analysis of it can make the most of the Long Data.
If anything, the process of workforce analytics and the revelations of its data are like a long journey, with a variety of steps for each organization over a considerable amount of time. HR executives are eager to rush along the way, and often anxious to get to the "predictive analytics" without building a solid foundation.
But the foundational aspect of workforce analytics is crucial; it tells an organization the What of its HR journey, while advanced analytics can describe the Why and, eventually, even more advanced statistical modeling can chart the course for sustained success. In a Big Data world that threatens to swamp us with terabytes, petabytes, zettabytes of information, the proper deployment of workforce analytics is essential if HR is to leverage the insights of Long Data -- and cut through the noise and fog of the digital ocean.
Brian J. Kelly is a partner at Mercer, based in Philadelphia, and is the global leader of the Analytics & Planning Center of Expertise. Julia Howes is the Boston-based product line leader for Mercer Analytics and Planning COE and the product manager for Mercer Analytics.