Whether as an interviewer or interviewee, it is a well-known reality that the hiring process is fraught with inconsistency and inherent human bias.
The fact that we have not evolved this process is magnified by three key shifts:
- Given the volume of applicants in a global market, companies cannot properly evaluate all incoming job candidates, resulting in filtering processes that are biased and/or random.
- Most U.S. companies take several months to hire, losing out on top talent that is snatched off the market in days or weeks.
- The world of work has become more fluid and complex, making resumes, degrees, and GPAs much less predictive of performance in the dynamic jobs of today.
Finance and marketing are among the industries leveraging huge data sets to make more predictive decisions. So with data collection at historically unprecedented levels, it is time to transform candidate hiring and assessment as well.
Consider the classic pre-hire assessment: 100+ questions that seem repetitive to the test taker. Yet, to the assessment creator, this is the bare minimum required to make an accurate prediction about future performance, as each new data point decreases the likelihood of a mistake and increases the predictive power of the assessment.
Machine Learning is transforming how data is leveraged in this space by transforming a company’s hiring power as well as the candidate’s experience. As opposed to multiple choice tests, large amounts of valuable data can be gathered from sources like video interviews, coding challenges, and games. The top four benefits of using AI to find patterns in this data are:
- Consistency. Approximately 93% of Americans think they are in the top 50% of drivers. Similarly, people think they are great interviewers, when they are in fact, very inconsistent and biased. With Machine Learning on structured interviews or tasks, candidates are assessed on a consistent set of criteria. This also does away with subtle unconscious biases that occur in unstructured interviews, like similarity bias — if you and a candidate went to the same school or have the same hobbies, etc., you are much more likely to recommend them.
- Fairness. Machine Learning algorithms allow for the possibility of removing bias against protected classes (age, race, gender) systematically — something that is impossible to do with human evaluators. If a model is found to adversely impact certain groups in a statistically significant way, any traits tied to identifying that group can be removed from the model.
- More Accurate Prediction of Performance. As opposed to traditional assessments, which often predict personality traits that are assumed to be necessary for a given role, this new paradigm relies on the direct prediction of job performance. Companies can find candidates who display the same types of skills and characteristics as their top performers for a particular role.
- Efficiency. Above and beyond hiring accuracy, the use of AI makes hiring faster for companies while also being more convenient and accessible for candidates. For instance, one international hotel chain reports that their hiring process has gone from 45 days to 5 days since implementing AI-scored video interviewing and Unilever has shrunk its hiring time from over one month to two weeks using games and video interviews.
The predictive power of AI holds huge promise to make hiring more pleasant and efficient. But unlike the finance and marketing industries, talent acquisition has a duty to ensure that the results of their analysis have no adverse impact on protected groups.
Understanding why and how an algorithm arrives at its conclusions is necessary to making changes that mitigate such adverse impact.
Already at work in companies around the globe, AI is disrupting the traditional hiring process. Enormous gains in efficiency and candidate experience are putting companies who do not adapt at a disadvantage when it comes to finding top talent.
Let’s block ads! (Why?)