Why Big Data in HR Needs Context, Stories and Pictures
July 29, 2016
Is a picture worth a thousand words... or a thousand big data reports? As big data becomes more prevalent in everyday work, studies show that HR leaders, analysts and their constituents must not only be adept at data analysis, but also at providing context, imagery, mental models and stories around this analysis.
When HR applications reside on personal devices, workers, managers and leaders encounter the findings of data more frequently—whether it's insight into their personal performance or findings about their team's productivity. This raises the bar for HR leaders to make data more engaging and compelling for the everyday user; a baffling array of "HR speak" or statistics—even on a personal device—will only turn away those who should be using it.
Here's what HR leaders need to know to make data more approachable in the workplace.
"Timing is everything" is evolving to "context is everything" in our lives. With smart machines, the Internet of Things and cloud data everywhere, "killer context" replaces "killer apps." In a connected home, for example, cameras recognize the faces of different occupants and adjust temperature and lighting to their preferences. On our phones, digital assistants like Now on Tap read over text messages to suggest dinner at a nearby restaurant or request to pick up this week's dry cleaning.
HR data needs to catch up. Microsoft's recent acquisition of LinkedIn has prompted speculation that an individual's personal and social information will provide context to Microsoft's Office applications. Moving forward, a calendar app may sense an upcoming meeting between a manager and employee, providing the manager with analysis about the employee's turnover risk, career aspirations, mentorship progress and crowd-sourced performance, with reminders of the employee's recent significant life events.
The Power of Pictures
Pictures make sense of data: Consider a scatterplot with a trend line, compared to a statistical correlation or regression equation. The scatterplot is more engaging. A recent study of 650,000 biomedical research papers used an algorithm to identify the use of graphics and found that papers with diagrams were more frequently cited; more equations, on the other hand, had reduced citations.
Leaders can similarly simplify big data in HR with diagrams and pictures. One CHRO repeatedly documented to her leadership team the statistics showing a lack of global diversity among organizational leaders. Yet, her leadership acted only after she presented pictures of the executives who had been consistently passed over for years.
Stories Tap Mental Models
For example, when HR analysts frame talent development as "risk optimization" or frame pay and benefits as "product design," they "retool" HR and invite their audience to use more compelling mental models than they might use if provided only statistics. They tell better "stories" about the data.
IBM used this method when they retooled workforce sourcing, career movement and attrition as a "talent supply chain" that was shared with managers and employees. The system showed what jobs and capabilities were in high demand and the development paths needed to qualify. So, when employees and their managers had performance and development discussions, employees could ask, "How can I become qualified for those high-demand jobs?" The supply-chain analogy plus the context of a performance and career discussion created a compelling data moment. Imagine taking this one step further: HR software that alerts both the manager and employee to schedule a conversation when a career opportunity arises.
For decades, the HR profession has strived to improve data quality and precision, statistical analysis and delivery systems. Analytics groups are populated with technical wizards trained in sophisticated disciplines like physics, psychology, sociology and statistics. But the most savvy HR organizations should also embrace writers, film makers and storytellers. Big data works when it inspires useful action, and that requires not only "analytics," but also context, imagery and a story.