Using AI to Drive Outcomes in Hiring and Leadership Development

For those of you that have followed our work over the decades, it won’t be surprising to hear our perspective on AI. Since our first SHRM published book in 2008, we have always emphasized the importance of driving actual impact (real utility) with any HR innovation. Numerous HR fads that promised to be game changers have come along (e.g. employee engagement, ‘skills’, the death of performance management, etc.) and never really delivered on their promises. The latest is Artificial Intelligence, which certainly promises the world (well beyond HR) to every leader who will listen. But before we make this the latest must-have—it is critical that we create a foundation of what we are going to use AI for when it comes to Hiring & Leadership Development assessments—and what criteria we should use to measure its success (or lack thereof).

AI in the context of assessments MUST first have a foundation of two important factors—proven impact on actual business outcomes (e.g. performance, turnover, financial, etc.) and no bias nor discrimination. Unfortunately, in the HR assessment vendor landscape, the trend is to just include AI in all the marketing materials and hope that no one asks any questions about impact and validity. Without the impact and validity verification, AI is just helping HR to get better at a bad game—namely to make more biased and/or non-impactful decisions at a quicker and more automated pace. If you are using AI based on a foundation of faulty algorithms, then you will get faulty outcomes—it’s as simple as that.

Making a hiring or promotion decision is one way for AI to have an impact in the organization—which is full of risks if not based on valid assessments. Another important way is for AI to synthesize the information from numerous assessments into a coherent, easy-to-understand and (most importantly) actionable story. Again, the foundation is what matters here—many organizations have fallen into the trap of thinking that a very short assessment of personality will tell them everything they need to know about a leader or job candidate. This, of course, is not possible from either a validity or bias perspective—so the most important message is to use an assessment that measures numerous facets (e.g. situational judgment, personality, working memory) and has proven validity and no adverse impact. Where AI can play a critical role is in synthesizing all the assessment information so that the leader and/or hiring manager and really anyone can make actual sense of what the assessments are saying and articulate it so that it is usable. To make it simple:

Assessments with Proven Validity/Impact based on Advanced Analytics + AI = Synthesized & Actionable Information that Drive Actual Business Outcomes

Imagine one of your leaders getting a comprehensive leadership development report and instead of them having to sift through a large number of pages to try and figure out what the data is saying and what they should work on—what they get is a summary of what the data is saying and what actionable steps they should take based actions that are proven to drive actual business outcomes (and there’s no inherent bias either). Integrating and summarizing the data is something companies pay organizational psychologies thousands of dollars to do for each assessment – a well-trained AI can do this in seconds at a significantly reduced cost.

It’s the same approach for hiring (leader or non-leader) into your organization. Getting not only an unbiased and highly valid assessments which lead to the best hire, but a synthesis of all of that information for the hiring manager to know how to manage that new employee for peak retention and performance on day 1 is an extremely valuable use of AI. The hired employee could also get a synthesized report using AI to help them take the right actions on day 1 of their employment to maximize performance. So, just because a tool uses AI, don’t assume that means it’s better – focus on AI that adds value or utility without adding bias in decision making.