Many of us who have operated within the human resources profession or been involved in strategic initiatives aimed at placing the workforce at the center of competitive advantage (aka human capital management endeavors), thought we were at least conversational about predictive HCM tools. We were aware that industrial and organizational psychologists have, for decades, been creating skill- and personality-based assessments using predictive algorithms that stood up to rigorous testing, and how tools such as the Hogan or Myers-Briggs tests became industry standards.
These methods proved quite impactful in determining a candidate’s fit for a job, corporate culture or team, or foreseeing one’s inclination to stay with an organization and perform well given certain conditions. Over time, these tools were joined in the predictive HCM arena by algorithms powered by other sources and types of data such as work history or how situational judgment or basic questions were answered. Predictive algorithms in HCM have been increasingly used to address which candidates should be green-lighted for interviews based on their applications, and which workers were flight risks.
While these capabilities delivered strategic HCM value for many years – and still do – two key questions emerged: Were they more of an expensive luxury for some organizations? Did their predictive value vary based on unique organizational contexts? These factors, combined with the explosion of people data housed within organizations, created a situation where technological advances had to occur before these tools could more cost-effectively predict workforce outcomes at scale.
And those technical achievements have been realized: Advances in applying artificial intelligence to workforce issues have occurred in earnest, albeit accompanied by excitement, confusion and/or skepticism. Today’s AI in HCM capabilities include the long-standing algorithmic techniques that guide HCM systems and tools in scoring the likelihood of different outcomes such as job success. They also include the use of natural language processing to enable workers to engage virtual assistants for quick answers without bogging down HR teams.
The confusion seems to mostly reside in relation to the machine learning subset of AI capabilities, which are arguably still nascent. Even the most powerful machines take considerable time and huge amounts of data to reliably detect patterns, learn from the patterns, and prescribe the best actions revealed by the patterns. The intense focus across the HCM vendor landscape on weaving a compelling AI story around their products has resulted in many buyers being more confused than ever about the AI capabilities in HCM that might have the biggest impact on their organization’s HCM objectives and priorities, where – or pursuing which use cases – would that impact most be felt, and whether the capabilities were as reliable as billed.
Upon joining Ventana Research in late 2017, I set out to convey what I had learned over the years about predictive HCM, beginning with this 30-minute on-demand webinar. By the end of 2018, and as the ML aspects of AI in HCM started to capture the imagination of organizations and HCM vendors, I started highlighting some of the capabilities one could expect, some of which are now upon us. These include the ability to personalize learning, making it more meaningful and effective, and doing so at scale – not just in relation to content, but to pace and medium. Complementing these explorations were my annual assertions or predictions for the HCM software and tools market that have increasingly featured the AI in HCM theme.
To help organizations understand which AI in HCM capabilities were real, ready and relevant, I created a use-case framework to serve as a communication tool in explaining why there’s so much to be excited about in this area, but also still so much value to be delivered. Practical use cases have always been my primary way to break through the clutter and help bring clarity and consistency to a potentially confusing HR technology topic. A more detailed version of the framework below is being employed by various HCM vendors and customers in their efforts to prioritize initiatives and communicate more clearly around this critical but still-emerging theme.
Five categories of AI in HCM use cases with indicative examples:
To apply the use-case framework in your organization:
As we know, a major component of effective change management is proactive and targeted communication. Those affected may give the impression they don’t support or embrace changes (such as technology changes) because they don’t fully understand them or how those changes impact them personally. Simple frameworks can certainly help.