What if we killed the job interview?

What’s your top strength? What are your biggest weaknesses? Where do you see yourself in five years? What makes you want to work for us? Who would play you in a movie?

Most of us have experienced the futility of job interviews, or at least had to answer some of these dreadful questions. Yet it’s virtually impossible to get a job without going through an interview–and usually more than one.

But the trouble is that interviews aren’t as useful as employers think. Indeed, organizations can still make great (and arguably better) hiring decisions without them. What would happen if we all agreed to scrap job interviews tomorrow, and focused instead on other indicators of career potential? Unthinkable as it may sound, there are at least three important data points that suggest replacing interviews with other, more predictive measures is the way to go.

Job interviews have limited value for predicting performance

The most comprehensive scientific study to date on the predictive power of different recruitment tools suggests that the typical job interview provides very little valuable information over and above psychometric tests, which tend to be both quicker and cheaper to administer.

For example, once you know a candidate’s score on a test of general learning ability, a typical job interview will only improve your ability to predict their performance in a given role by 4%, the analysis found. Interviews are more useful when they’re totally structured and standardized, to the point of resembling a multiple-choice questionnaire; this can increase their accuracy by up to 13%. Yet very few real-world interviews follow a rigorous format. Interviewers usually prefer to go with the flow, stubbornly relying on their own intuition.

Most of the attributes interviewers try to evaluate by gut feel–a candidate’s competencies, skills, personality, values, “culture fit,” and so on–are more rigorously inferred from other data like resumes, simulations, tests, and past performance ratings. Interviews certainly create opportunities for candidates to make claims about these qualities, but as I argue in my latest book, there’s little reason to believe them. Indeed, there’s not much overlap between the talents people say they have and the ones they actually possess. (Plus, interviewers often use the idea of “good culture fit” to justify hiring people from their own in-groups.)

Job interviews introduce toxic data

In addition to the information interviews should provide but don’t, there’s also a great deal of information they shouldn’t provide but do. The latter isn’t just “noisy” data in the sense of not improving predictiveness–it’s actually toxic, focusing interviewers’ attention on problematic traits. For example, it’s all but impossible to ignore (and make biased, misguided assumptions about) a candidate’s gender, age, race, appearance, or social class, even when the most conscientious recruiter or hiring manager strives to prevent these factors from influencing her decision making. In fact, the more we try to ignore these qualities, the more present they’ll be in our minds.

Some technological solutions, including AI-based tools that scrape interview footage and translate what people say and do into predictive performance scores (among other tactics), are quite promising. Unlike humans, computers can learn to systematically ignore candidates’ ethnicity or gender while still focusing on relevant indicators of potential. And for all our fascination with unconscious bias, when it comes factors like race, gender, or nationality, people are quite conscious about their biases–sometimes they’re even proud of them. Worse still, interviewers who are aware of their biases will be better able to mask them, attributing their biased decisions to other, more legitimate factors.

Interview performance and job performance are two very different things

All interviews are essentially artificial social situations in which candidates are asked to present the best version of themselves, all while pretending to be themselves. In other words, they’re exercises in “impression management,” and there are big, individual differences in how well people handle that. The real question, though, is why we care so much about a person’s ability to display a desirable behavioral repertoire during an hour or so. Unless they’re being considered for a sales or customer service job, it’s just not a terribly relevant skill set. To be sure, political skills–and even acting skills, to a degree–can be helpful career lubricants, but they don’t say much about performance on the job.

In fact, so-called “dark side” personality traits, such as narcissism and psychopathy, are found among people with otherwise strong social skills, at least in short-term interactions, which makes them perform rather well on interviews. In that sense, interviews are just like a first date: Just because someone charms you the first time you meet them doesn’t mean you should marry them. The regrettable fact is that there are parasitic people in just about every organization–those who climb the ladder while sucking up resources and taking credit for others’ work, all at the expense of strong performers who go unrecognized and stagnate in their careers. De-emphasizing job interviews–or ditching them completely–might help alleviate this this problem.

Of course, that’s unlikely to happen anytime soon, and it’s important to acknowledge job interviews’ value alongside the drawbacks. For one thing, they’re great opportunities for candidates to learn about employers; research has shown that candidates generally like interviews and find them fair. However, it’s entirely possible to remove the vetting components of the interview and just let candidates have an informal chat to learn about the opportunity without being evaluated themselves. If not, their questions for interviewers will be just another means of impression management and performance display. (If a candidate candidly asked what they really wanted to know, they might not get the job.)

But the main reason job interviews are likely to stick around is that it’s hard for employers and recruiters to accept that their natural talents for identifying potential aren’t as strong as they may think. Yes, some people are extremely good at interviewing and evaluating others’ potential suitability for a job, but the vast majority overestimate their abilities on that front. Which, after all, is only human.