How HR is Becoming a Science

HR is becoming a proper science.

Between the newfound ability to connect HR initiatives to business outcomes and the explosion of intelligent tools, technology is making it increasingly possible to understand and predict the organization and its behavior. We’re still in the early stages, yet it’s easy to imagine an evolution in which HR can help business leaders anticipate the results of their people-related decisions and guide them through complex decision-making.

The spectrum of improved tools include reporting, people analytics, predictive analytics, machine learning and natural-language processing. Coupled with burgeoning frameworks for understanding work, people and organizational cultures, these tools constitute the exploratory elements of the new science.

From here forward, the hallmark of HR will be its ability to use data to steer the workforce towards the organization’s future.

The goal of all science is to understand the world in ways that make it more predictable. Until now, organizations and their members posed impossible challenges in modeling and predicting behavior. Even the best 20th-century assessment tools left much to be desired and offered little in the way of useful correlation.

Both people and organizations are examples of complex systems. They behave differently in different contexts, and minor nuances can generate extremely different outputs. The combination of data models, inexpensive computer processing and virtually limitless storage in the cloud make what used to be impossible calculations now possible.

In the beginning of the emergence of intelligence tolls (and that wasn’t very long ago), it looked like they would be added to core software processes like apps on a smartphone. There are over 600 venture-financed experiments involving tiny, unconnected bits of HR–a good model for exploring the limits of possibility and a bad one for integration and connectivity. So, what’s happening now is that little start-ups are essentially in a tryout stage for absorption into a larger entity.

Every year, I create a report that describes the issues and players in the emerging science of HR. My presentations at this October’s HR Technology Conference in Las Vegas are the results of hundreds of hours of interviews, demos and a deep look at machine learning, AI and how they work. As I research, I’m starting to see an uptick in absorption of these smaller experiments; the big fish are feasting on the little ones.

The danger is that the acquiring operations rarely produce a cohesive big picture. If the little, focused AI apps run the risk of unintended consequences, an assembly of them can multiply the problem. Whether you like it or not, you are going to spend a considerable chunk of your time figuring out how to dummy-proof your systems from the AI that has come to live there.

It’s getting clear that any large organization (say, more than 5,000 employees) will want to have its own analytics and data-science laboratory. That’s where organization-specific nuances and algorithms will be built. Here’s the start of a reading list that will get you in shape for this new, scientific HR:

• Organizational Culture and Leadership, fifth edition, by Edgar Schein: The classic graduate overview of organizational culture is a readable, fantastic introduction to the science.
• The Model Thinker: What You Need to Know to Make Data Work for You, by Scott Page: This is a solid review of 35 basic statistical models that can be used to simulate individual and organizational behavior. A key point is that system behavior is not the aggregate of individual transactions. There is a related online MOOC that is worth the 40 hours it requires. Google it.
• Superforecasting: The Art and Science of Prediction, by Phillip Tetlock: As HR becomes a predictive science, this is a manual for developing better forecasting skills. Most of the important areas of HR involve predicting future behavior as a result of present actions and interventions. This is a good starting point to understand and develop the skills needed to become an HR scientist.
• The Human Use of Human Beings, by Norbert Weiner: This is the first book on the ethics of intelligent systems. The various human risks associated with intelligent tools are not always obvious. Weiner was one of the first thinkers to delve into this area of AI, and his cautions are worth understanding.