Take This One Weird Quiz and We’ll Tell You If We’ll Hire You
People don’t really like to talk about how they got their jobs. If you ask strangers, you’ll see what I mean. What comes up often is “I knew someone.” Sometimes a teacher or a former boss: more often a relative, or a close friend. When they say this, people are actually answering the question: “How did you know a job was available?” They do not answer the question: “Why were you hired, instead of other applicants who were not?” Often people can’t tell you, because they don’t know. The job interview has become longer, larger and harder than ever before. Across industries, positions and the country, companies are scheduling more interviews for applicants to pass and setting up complex tests and assessments in an attempt to gauge not only job seekers’ competency but also their personalities. The result is a burdensome, frustrating process for the job seeker and unclear benefits for management.
Lots of people have a memorable job interview story — a time that they were embarrassed, or wore the wrong shoes, or gave great answers and got offered the job on the spot, or gave okay answers and then came home and realized their fly was open the entire time. It’s difficult to tell a good story about a time you took a test online: I know, because I tried to do it several times in this article. And it’s hard to learn anything about a potential place of employment from taking the skills and personality assessments a vendor sold them. In the headlong rush to find something, anything, that offers a full-time paycheck and a little insulation from the insurance marketplace exchanges, assessments are becoming a normal feature of the job interview process.
Hiring managers say that they are frustrated with the large numbers of applicants they get for every single job, since it’s so easy to just click ‘Apply.’ Predictably, job seekers are frustrated with the large number of jobs they have to find and apply for, since the response rate is so low. Often, assessments aren’t reserved for the most eligible candidates, or the candidates whose applications contained key words, or those who have already passed an interview or phone call with the hiring manager (or a phone call with an automated system). They’re part of the initial application, just like a resume and a cover letter. It’s not unusual to read a job posting, follow the instructions to send in a resume and cover letter, both written for the specific job, next fill out a form reiterating the contents of the resume and cover letter, and then receive an email with a link to an assessment, stating your application is not complete until you fill out the assessment. It may be short, lasting only ten minutes, or as long as half a day. Hiring managers want to talk to motivated candidates who understand the job and the company. But the reward for all those detailed applications and assessments is…silence. They add up to a lot of unpaid hours trying to give businesses what they want, even when they don’t know what they want or what they want isn’t humanly possible.
One medium-sized company I interviewed with this past summer asked me to come in four times for interviews, as well as separately take a language exam, take a math exam, and drive out of town to pee in a cup for a drug test, which cost me a fee on top of the gas money. This was for a part-time job. They sent me a form email telling me I didn’t get the job after all of this, a full two months after I first applied. There’s no way this labor-intensive hiring strategy can be efficient for that company. Or any company. That’s over a day of working hours spent interviewing each applicant. It’s tough, if not impossible, to schedule all those interviews and requirements without alerting your current employer to the fact that you’re out there looking and missing work, or if you’re unemployed, protecting your dwindling savings account or whatever interim gigs you were able to find.
Then there’s the cost for the assessments. Sometimes the job applicant pays for them, but more often the company does. Training someone for a job costs money and takes time, and it can take a lot of money, and a lot of time. But so does testing multiple applicants, and paying for their tests, and spending the time required to schedule and conduct multiple separate interviews with each one. Is part of the message “It’s tough to get a job, so accept what we offer you and don’t leave until we tell you to?” Who looks around their kitchen or lab or office or shop and says “We need extra hands, but maybe in like three months and only if they pass a personality test?” Apparently, more and more people.
The job assessment industry is thriving, with startups and well-established players. Kenexa, an IBM company, apparently lets customers choose from 1,200 different assessments. SelectiveHire offers industry-specific assessments, assessments that claim to measure task performance, and something labeled an “Integrity” test. Indeed, the job search site, has an entire wing dedicated to assessments. You can set up a phone call with a recorded voice to ask people classic interview questions (“What accomplishments are you most proud of? And why? BEEP”) which you then record and listen to later. Or you can set every applicant a multiple choice quiz where they read a schedule separated into blocks and pick which time and day John Executive, and it’s always a John, can schedule an appointment. For some reason most of the automated phone calls are scheduled, which means you planned ahead and made time to make a recording for the hiring manager to listen to when it was most convenient. Ziprecruiter’s ads describe hiring as an arduous and terrible process, one that their service will solve for you via mysteriously effective tools. In one spot played on a podcast, they tell the story of a woman who posted an ad looking to hire an animator. She picked one who spent three hours writing an answer to her question about why education mattered to him, the three hours of course tracked by Ziprecruiter’s software.
At least she actually read the answers, which appears to be a rare practice. You can’t ask an employee to stay three hours late without pay, but a job seeker should do it to show motivation.
For all the talk about using cold hard data as an improvement on the fuddy-duddy old method of reading resumes and calling references, and the new age of information, it’s hard to find any reliable information on whether these assessments are good predictors of later job performance. When I applied to a researcher job opening at a three-person financial services company, they asked me for my ACT score. To proceed, I had to pass the hiring agency’s skill assessments, too, which were passed along with my resume, cover letter and ACT score from when I graduated from high school seven years ago. The company interviewer I met spent most of the time rambling about how great his business was and how much it impressed people, pausing briefly to ask me where I went to high school and what my dad does for a living and then cut off my answers. He told the hiring agency I was bright but wouldn’t be a good fit for being “too academic.” I fell off the tightrope stretched between having a good academic record while appearing uninterested in the company. Still, the ACT score question revealed an angle I hadn’t previously considered: the job interview test as a continuation of the standardized testing that runs the US public school system and plays a large role in childhood and adolescence. Like those tests, part of what these long, drawn-out job interviews measure is the willingness to follow directions.
In the course of an internship application, I ran into a test from the startup Pymetrics, which assesses applicants’ personalities based on a half-hour of internet gameplay. Yes, you play games, clicking on bars as many times as you can for sixty seconds or pressing the right arrow for a green sign and the left arrow for a red sign, and then the site pops up with little sentences that tell you about how you assess risk and make decisions. “You have a generous streak that isn’t affected by the situation at hand.” Yes, I’ve donated my time to playing this game. They reminded me of a CAPTCHA, those little tests that ask you to pick which images have a traffic light in them or type out a word in a strange font, except they’re longer and are supposed to help you prove you deserve healthcare, not to get on a mailing list. I never heard back about the internship. But I did get four separate emails from Pymetrics asking me to fill out a survey about my experience to help Pymetrics improve. The company spent money on this assessment software, but not on paying interns.
On its site, Pymetrics describes itself as a “gamified recruiting tool powered by AI” and touts its potential to remove selection bias from the hiring process by comparing applicants’ scores on neuroscience-based games with existing employees’ performance to find the best match. As the saying goes, a neuroscience-based game is as good for the goose as the gander, but only if the goose and gander had equal amounts of time and energy to get really good at video games and the hand-eye coordination required, even if the directions imply that the number of times you can click on a target within sixty seconds has something to do with your mental or emotional capabilities or what you have to offer as a professional. Therapists who have worked with the same person for years on end find it difficult to describe their patients’ tendencies and capabilities or predict their behavior, but clicks can do it in 30 seconds.
The websites for services aimed at hiring companies emphasize how they can filter an overwhelming number of applications. The estimates for the number of applications the average job posting receives vary widely: 250? 75? 300? 500? Whatever it is, it’s an overwhelming burden, one no number of salaried human resources employees, no matter how hardworking, efficient and motivated, could possibly handle. It also represents an astonishing amount of other people’s time and energy spent searching out job postings and preparing applications. If employees are really as important to employers as employers say they are, most of them have a very funny way of showing it.
There is also an entire YouTube genre of ‘How to Pass an Employment Test’ in Microsoft Office, Excel, Salesforce, whatever you can think of. Commenters trade tips, ask for good luck wishes, and thank video posters. There are only very rarely hints about what will be on the tests, meaning a job seeker can’t help showing up unprepared. For one person applying for one job, taking a test isn’t such a big deal. Multiplied across many, many opportunities? They add up. They aren’t a final step across the gate, either: in most cases the applicant never receives a confirmation that human eyes have seen their resume and cover letter, however briefly. People best placed to pass the test are those with Internet access at home and pretty good computer skills already.
An assessment sends the message that we will not train and experience is required. There are right and wrong answers. I took an Excel test as part of a four-hour interviewer for a part-time email customer service job at a relocation services company in February. One interviewer had told me, unprompted, that I was the first applicant for the job to be interviewed, but when I saved the completed test to the desktop as instructed, I found an identical test with a man’s name on it, dated from the day before. I didn’t look at it for an answer about how he managed the pivot table in part six, even though it was tempting. They emailed me three weeks later to say it was a full-time job now and ask if I was still interested. I told them I was and then I never heard from them again.
The skills-based assessment can be justified on the grounds that it can help to establish which candidates have skills relevant to the job. The personality assessment has no such credibility. There are lots of problems with the popular Myers-Briggs Type Indicator testing, some of the most pressing being that it very rarely gives consistent results one day to another and that it sorts people into polarized extreme types. Completing a thirty-minute assessment of your feelings, goals, dreams, relationships, and bad habits for an entry-level administrative job feels sort of like being asked to get naked and talk about your childhood fears at a hair salon so the stylist can see if the trim you want would suit you. There’s no guarantee that someone who has a similar personality to other coworkers, even if a test could measure such a thing, which it can’t, will be a better employee than someone with a different personality. Working people, it turns out, must be protected from the great unwashed population of those whose personalities do not match theirs when tested. How they manage to communicate with customers is a mystery.
Perhaps a company needs a new employee who is stubborn enough to insist that every single safety rule be followed, or someone antisocial enough to refuse to participate in workplace gossip, or someone argumentative enough to break old and unproductive patterns. The personality test as a resume filter explicitly prioritizes personality over skill, intelligence and motivation. What they want to know is probably, “How will you handle the problems of this job? What would you do in this job if this specific situation arose?” What they ask is…confusing. No job seeker ever receives the scores of existing employees. No job seeker ever receives their own scores.
Whether from a fear of being sued, shyness, or the inability to explain themselves, companies do not give job seekers, even those they’ve assessed and interviewed and those they’ve interviewed multiple times, any feedback. There is no way to tell that if you spoke another language or learned more Excel skills or simply held on to your current position for another year, you would be successful next time. Significantly, there is also no way to tell that the job was earmarked for somebody’s cousin or your last name didn’t seem right.
The personality assessment allows the company unlimited and unquestioned refusal on the grounds of ‘fit.’ There is no wrong or right candidate or quality, just a good or bad fit, just some people who got hired and many people who didn’t. The language of ‘fit’ is awfully convenient. Only companies determine ‘fit:’ job seekers can only guess at it. Fit, basically, means “We don’t want to tell you.” It may also mean: “We can’t.”
I agree with the testing assessment companies’ pitches in one spot: the hiring process is broken. I have avoided using the term ‘system’ throughout this article, even though it appears in what literature exists about the topic, because ‘system’ implies a series of events with some degree of predictability and consistency, which is a far cry from anything I’m talking about here. The whole process is as opaque, mysterious and vague as a medieval trial by ordeal. This is partly because companies won’t give away their secrets and partly because people aren’t honest about how they got their jobs. Any openness would end the polite fiction of “We’re looking for the best match in terms of skills.” It is a massive waste of everybody’s time and resources, it is infested with deeply rooted nepotism and prejudice of all kinds, and it is ineffective at matching people with jobs that fit their skills and needs.
Everybody knows that the labor market in the US is tight right now, with unemployment down and job postings up. In context, removing barriers between applicants and open jobs, rather than seeking out new and “data-powered” ways to create new and “data-powered” barriers. As tempting as they seem, tests are not the answer to anyone’s hiring problems. The communications gap will not be filled by video interviews assessed by AI, or multiple choice questions, or Myers-Briggs testing. There are solutions that may help: Job sites could limit the number of jobs seekers can apply to in a week or a month, forcing them to concentrate on only those most relevant and realistic to them and giving employers manageable numbers in return; or it could limit the number of applications a posting can take. Employers could (deep breath) pay the most promising applicants for a half-day of testing. Job applicants and job posters alike could get honest and spell out what they mean by ‘organized and positive’ and ‘intermediate Excel skills.’ This kind of effort requires coordination, and change from existing postures, and some kind of recognition of a less-is-more, quality-over-quantity philosophy. In other words, a stretch.
