IBM is using software to solve HR problems, and created a new consulting practice for organizational changes: http://t.co/LGpWCjGjZH
— Mohana Ravindranath (@ravindranize) August 7, 2014
I found the wording of this tweet and article summary more than a little interesting: What kinds of “HR problems” and “organizational changes” will IBM “solve” with analytic software?
From the Washington Post article (emphasis added):
[IBM] unveiled several new technology services that would apply big data and analytics processes to human resources problems.
One service, predictive hiring, would use large volumes of behavioral assessments and other employee data to better understand the traits that are characteristic of top performers, and then comb through candidates to identify potential hires. A predictive retention service would analyze workforce data — exit interviews, for instance — to identify those employees most likely to leave.
What strikes me here is a seemingly implicit assumption that employee “traits” are largely constant, innate, “hard-wired.” Either people bring the traits the company believes are those of high performers (or job hoppers) with them through the door, or they don’t. Identify who has (and doesn’t have) those subtle traits, and you can hire and keep top performers while winnowing out the others.
Color me skeptical, both of that apparent assumption and whether big data analytics is even the right tool here. I am reminded of Maslow’s law of the instrument: If what you have is a hammer (in this case, an analytic hammer), every (HR) problem will look like a nail.
First, the analytics themselves described in the article seem to be at high risk for falling victim to logical flaws described by David McRaney in his must-read blog-turned-book You Are Not So Smart.
For example, since existing leaders will largely decide what those attributes of top performers are that require big data and predictive analytics, they might well end up identifying ideal employee profiles in their own image (confirmation bias).
And analyzing the traits of perceived top performers with a long history with the company appears to run afoul of survivorship bias:
You must remind yourself that when you start to pick apart winners and losers, successes and failures, the living and dead, that by paying attention to one side of that equation you are always neglecting the other…
[For example,] when a company performs a survey about job satisfaction the only people who can fill out that survey are people who still work at the company. Everyone who might have quit out of dissatisfaction is no longer around to explain why. Such data mining fails to capture the only thing it is designed to measure…
Second, human workplace attitudes can and do change significantly over time, particularly in response to the organization’s own traits; for example, whether the work environment is risk averse or proactive, collaborative or politically charged, collective or exclusive. And employees don’t need big data or predictive analytics to draw those conclusions.
Liz Ryan, former HR VP and founder of Human Workplace, hit that nail right on the head (bad pun intended, sorry not sorry) in this recent article:
In order to hit our goals in any organization, we need to build positive energy in the workplace. We need people to be excited about their work… Anyone in your organization will be able to tell when the excitement level is high… low, or nonexistent…
Physicists proved a hundred years ago that energy moves in waves, and we see waves around us in the air and water. Still, we pretend that the workplace is a linear place made only of easy-to-measure particles… as though the waves aren’t there.
To her credit, Liz Ryan doesn’t pull any punches sharing her disdain for technocratic measurement of employee engagement (and by extension, performance) instead of other much more effective ‘low tech’ methods:
If we really care what our employees think, it’s easy enough to find out… We could ask them how they’re doing… We all have enough creativity and intelligence to move our organizations without unnecessary, fear-fueled micro-management practices…
The more formal, rigid and hierarchical an organization, the less easily waves of energy and trust will flow. If the energy is moving, problems will get aired and resolved quickly. Trust builds trust. Trusting people will hire other trusting people, and everyone will win.
While Liz Ryan’s key point is right on target, I do think she takes her criticism of measurement-driven management too far. Business intelligence and analytics, when created and used appropriately, are a powerful force multiplier. But analytics are not the only arrow in the technology quiver. Other tools have been specifically designed to actively contribute to employee engagement and an “espirit de corps” which in turn drives new employee productivity.
Gamification platforms is one such tool I have become aware of right here in Boston; for example, Objective Logistics‘ MUSE platform engages and motivates restaurant workers, while the WeSpire platform engages and energizes employees around company sustainability and social responsibility programs (of course, both platforms are complemented with analytics).
Making a conscious, personal effort to build an energetic, collaborative work culture should yield much better employee outcomes than poring over predictive analytics based on the past and potentially flawed assumptions. “The best way to predict the future is to create it” is an old saw, but one that still rings true when considering the right actions and technologies to improve employee performance and successful hiring.