A July article in the Chicago Tribune (http://www.chicagotribune.com/business/ct-biz-artificial-intelligence-hiring-20180719-story.html) chronicles the demise of resumes in hiring, and the rise of chatbots, artificial intelligence, automated games that “reveal” traits, and video interviews that measure 20,000 facial expressions (HireVue). The goal of all these approaches is to save money (fewer interviewers) and minimize bias. Proponents see some reduction in turnover. Problems with the approaches include MORE bias when algorithms try to repeat patterns of success and successful employees happen to be old white guys from Harvard. Candidates complained of auto video interviews in which follow up questions asked by the canned voice did not relate to the previous answer. Personality tests are most common – a $4 billion industry. Companies just sent a link, candidates take the online test, and those who score below a cutoff score are rejected. Simple! But … do they work?
Brad Smart’s opinion: Yes and no. Yes, I can see that technology will help improve talent through better hiring tools. But no, they are not “there” yet. The CEO of a major HR technology company told me there are 40,000 apps in the HR technology space, all promising to improve your hiring but NONE of the new apps showing significant improvement in talent.
I happen to be an expert on personality tests, having properly validated personality tests. Every personality test with a cutoff score that I’ve validated did more HARM than good – it eliminated as many A Players as C Players. (Maybe I should do a 15-minute webinar and prove it). YOU can do a short version of what I’ve done with 10,000 candidates over a 3-year period.
- Administer your personality test to at least 30 candidates (BEFORE they are hired), and don’t score it.
- After 6 months bring out the test. Score it. Get performance ratings on all (12) who were hired.
- Count how many high performers (A Players) were rejected (not hired because their score was below the cutoff).
OOPS! If as many As as Cs were rejected, the test failed itself!
I have my own built-in bias – in favor of the Topgrading approaches with proof of improving talent in dozens of case studies (www.topgradingcasestudies.com). I predict that technology will eventually help improve the efficiency of Topgrading, but I also predict that:
- resumes will prevail (though right now resumes and Facebook profiles and LinkedIn profiles are loaded with fabrications), and I predict that …
- when candidates know they will have to arrange calls with bosses the low performers (with falsehoods in their resume) will drop out, saving companies time and money. And I predict that:
- good old fashioned face-to-face interviews will be the only way to fully, fully understand candidates, and finally, I predict that
- no-phone tag reference calls with managers of candidates will continue to be the solid verification of what candidates told you.
What do you think?