HR Strange But True

The Voice—What Algorithms Can Tell You About Applicants

Applicants may walk the walk, but now computers can determine if they can really talk the talk!

Yes, hiring sales and customer service staff may soon be getting easier—if you are willing to let computerized algorithms do the selection for you, according to an article on NPR’s All Tech Considered blog.

Sure, you may want humans in HR to prescreen applicants by skills and experience, but the algorithms will determine which applicants are likely to be more successful at convincing customers to buy your products or at handling an angry caller.

How? Their voice! The article says years of voice analysis and focus groups have resulted in the key elements in the human voice that categorize the emotions of a speaker. Human voices have biological-based markers, like fingerprints, that are unaltered by accent or age. This means the voices of Al Pacino and Meryl Streep could be identified no matter how old the movie or what accent they were channeling.

Luis Salazar, CEO of Jobaline, which researched and created the algorithms, says they do not gauge a person’s emotions or even if they are lying or telling the truth. “That’s irrelevant, he says.”

The workplace application? Salazar told NPR that the voice analysis can be an objective means to see which applicants’ voices listeners would find the most engaging, empathetic, calming, or trustworthy, attributes important to applicants for jobs in the healthcare, hospitality, or retail industries in sales or call center positions. Sounds like we will see more of this concept in the future.

HRSBT wonders what those algorithms would say about those screaming car salesmen on the late-night commercials!

1 thought on “The Voice—What Algorithms Can Tell You About Applicants”

  1. Sounds vaguely discriminatory. Plus, voice is just one factor that will determine an employee’s success in a particular position, and, of course, coaching can bring about improvement.

Leave a Reply

Your email address will not be published. Required fields are marked *