Kevin Parker, chairman and CEO of HireVue, recently shared with me his perspectives on the future of hiring. HireVue has created a platform for automated job interviews, combining video interviews of applicants with AI-powered algorithms that can sift through their qualifications faster than any HR team. Because 100% of applicants can complete a video interview on their own schedule, companies have access to a broader candidate pool. And because the algorithms do the initial screening, the process isn’t limited by the HR team’s staffing level.
Of course, no algorithm can evaluate candidates the same way a human would. Like any breakthrough in machine learning, it’s a grab bag of promising signals emerging from the noise and, well, noise. As Kevin will tell you, though, the noise-to-signal ratio is falling fast as HireVue gathers data and experience.
During our conversation, Kevin discusses whether AI-powered algorithms have the potential to automate résumés out of existence. We also talk about how the rules-based decisions of a machine may eliminate some human biases, while amplifying others. As with other AI or machine learning–based models, HireVue’s interview machine can develop rules for ranking candidates that would be nearly impossible for a human interviewer to deploy (counting the number of times a candidate uses a particular word, for example, or tracking eye movement). Because this interview experience is so radically new, it takes some getting used to. So we also talk a bit about how to acclimate both employers and job applicants to the new world of video interviews.
In the following excerpt, Kevin describes how his customers make the leap into an era of automated hiring.
Kevin Parker: The best way to think about HireVue is an on-demand platform for job interviewing. It allows candidates to take all the time and place and logistical challenges of doing an interview and put them aside, because they can interview for a job 24 hours a day, 365 days a year, on any mobile device and any laptop. So it’s a very democratizing view of the world, and we have customers that are interviewing a thousand people a day for jobs around the world. We do it in 32 languages in over 180 countries today, all on demand, and so it really has changed the way companies look for talent.
Rob Markey: So just to be clear, instead of arranging an interview at a specific time, on a specific day, with a specific interviewer, this allows me, as an applicant, to simply interview using my computer or my mobile device, have it recorded and then sent to the company?
Kevin Parker: That’s right. If you were going to take an interview, you would log in and the company would have set up some prerecorded questions—not an extensive interview, maybe five, six, seven questions—just to learn a little bit more about you and so you can tell a little bit of your story. The advantage is that, one, obviously, you don’t have to travel and you don’t have to take time off work. But you’re also getting a very consistent experience. You know that every candidate got asked exactly the same interview [questions] in exactly the same way, and so it’s a much fairer process. All the cognitive biases that come into interviewing a person are put to the side because everybody’s having the same experience.
Rob Markey: Now, one of the benefits of this is that it allows more people to apply for any given job. How do you deal with the fact that there’s still limited bandwidth for the interviewers to review all of those interviews and to go through all of the candidates?
Kevin Parker: That’s a really good question. And that’s one of the things that volume causes, you know. We’re interviewing very widely. We’re opening the funnel very wide in terms of talent, but we’ve developed artificial intelligence applications that allow companies to look for the very best talent in that pool of folks and move them to the front of the line. And so we’re using artificial intelligence to turn sort of a traditional, long, first-in, first-out process into an accelerated, best-in, first-out process.
Think about the typical hiring process. It may take 60 days to get all the way through, most of which is characterized by inactivity on either side, the candidate side and the recruiter side. Whereas if you can take an interview and I can score it using artificial intelligence, and if you’re a great candidate, the recruiter can reach out to you the next morning. So it really changes the dynamics of the process in a very different way and allows it to accelerate.
Rob Markey: I think people would be skeptical about this idea of having artificial intelligence substitute for human judgment. So help me understand how it is that artificial intelligence can provide a better solution than a human interacting with me as a candidate.
Kevin Parker: Well, that’s a great question. What we’re really trying to do is replicate in a very consistent and very scalable way what would happen when you were sitting across the table from an interviewer.
We have an IO [industrial organization] psychology team that will work with a customer to design a series of questions that are workplace related and skill related, such as “How do you deal with customers stuck in a difficult situation? Tell me about a time you worked together as a team, and the challenges and successes that came with that.”
So we’ll design questions almost as though we’re designing an interview guide, but present those in a video context, and then we’ll look at those answers and correlate them to success within the organization. We’re looking at the words that are used, the speech-to-text translation. We’re looking at the stress, we’re looking at the number of times you use the word “I” vs. the word “we.” And so we’re really trying to get at a very scalable level and replicate what humans do automatically, but also extraordinarily consistently. If you were sitting down for an interview on a Friday afternoon, you may be a great candidate, but if the interviewer is already thinking about the weekend or what they’re going to do tomorrow morning or, you know, their vacation plans, you’re not going to have a consistent experience.
That’s really at the heart of what we’re trying to do, to look at those answers and correlate those to performance inside the organization. So the best team members may be empathetic, they may use the word “we” more than “I,” they are team oriented, and we can look for those attributes in the interview at scale.