Brittni Daron jumped through a lot of hoops before she landed her job as a solution consultant at Oracle. At the tech giant, as at other firms in Silicon Valley to which she applied, she endured weeks — and occasionally months — of phone interviews, in-person interviews, mock presentations, personality tests and technical tests for both the skills she claimed to have and those she didn’t. This might sound a little ridiculous, but it’s not unusual. I’ve met lots of job seekers in the last few years who underwent a similar form of H.R. torture — spelling quizzes, math exams and oddball brainteasers (“If you were a pizza-delivery man, how would you benefit from scissors?”) — only to walk out of the interview room wondering what they’d really just been tested for. “Basically, I could never figure out what they got out of it that qualified or disqualified me for the job,” Daron told me.
These days, companies have become more cautious about whom they hire. The average length of the job-interview process has nearly doubled from 12 days in 2009 to 23 days in 2013, according to an analysis by Glassdoor, a website where users rate employers. (Curiously, biotech and pharmaceutical companies averaged the longest interviewing periods, at 29 days per candidate.) At the same time, many employers are also trying to become more cost-effective. Since interviewing takes time and resources, companies have increasingly focused on finding work-force-analytics software that can make the process cheaper and more efficient — and ideally better at identifying the most promising candidates early on. “There is a huge amount of money to be made in matching people to their careers better,” Erik Brynjolfsson, an economist at M.I.T., told me, “and improving the allocation of trillions of dollars of human capital.”
Deep Thoughts This Week
1. Companies are trying to hire more efficiently.
2. Which would make the overall economy more efficient.
3. If only we knew the right algorithm.
Brynjolfsson is advising one start-up that hopes to revolutionize what might be called the human-capital-allocation market. The company, Knack, uses video games as a screening device to determine how creative, cautious, adept at multitasking or easily distracted, among other attributes, potential job applicants are. Its “Wasabi Waiter” game, for example, casts the player as a server at a sushi restaurant who must figure out what dishes to recommend. It’s one of the more gimmicky-sounding entrants in an already crowded field competing for that “huge amount of money.” Other companies include start-ups like ConnectCubed (another gaming-based assessment technology), Good.Co, Evolv and Prophesy Sciences, as well as older firms that also provide other personnel-management tools, like Kronos.
These companies are an outgrowth of the “Moneyball”-style, buzzwordy “big data” empiricism that has transformed fields from professional baseball to political campaigns. H.R. firms are increasingly administering newfangled tests to existing employees to track which traits or habits correlate with whatever the firm considers to be evidence of success. Then they try to identify those characteristics in potential hires through cold, hard observational data, rather than through hiring managers’ far mushier gut reactions and candidates’ self-reported declarations of “greatest strengths and weaknesses.”
Watch Now: Inside China’s Internet addiction bootcamps
Watch: Urban grit in Philly’s Fishtown
Watch: How to make oatmeal sandwich cookies
The idea that a video game or a fancy test could, in one fell swoop, better and more cheaply match people to their jobs is a seductive one. But it’s also a bit trickier than finding a way to recreate Jason Giambi’s offensive output at a fraction of his salary. Administering a test or a game remotely has the potential to pull in a lot more job candidates who might not otherwise have been considered — people who didn’t go to elite schools, for example — but it could also screen out entire classes of workers. Millennials might be totally cool with playing a video game set in a sushi restaurant as part of a job application, but some older workers might balk. (Then again, even I, a millennial, got a little sweaty-palmed when I sat down to play Wasabi Waiter.)
There’s also the matter of the quality of the tools themselves. Do they really measure the things we want them to measure? A lot of these start-ups promote projects they’ve done with individual companies that show strong correlations between some test question or observable behavior and various success metrics. Knack has worked with Shell, for example, and a lot of medical employers; and Evolv with Xerox and Harte-Hanks. But in an economy where work processes and the requisite skills change very quickly, it’s not clear that yesterday’s correlation is a helpful predictor of whether an applicant will be a high performer tomorrow, said John Sullivan, an H.R. consultant and management professor at San Francisco State University. In the meantime, you might worry that job applicants can figure out how to game the system. Or that you’re just cloning the workers you already have, rather than bringing in new perspectives.
Interesting article. I was most interested, though, in a comment near the end: “Three-quarters of the nation’s wealth is in the form of…
Earl Hall 43 minutes ago
As a plaintiff’s side employment attorney I look forward to the latest “flavor of the week” consultant gimmick. Musabi Waiter, indeed.
LI’er 43 minutes ago
Great to read about this trend…? Only if such a process could be used to reach out to all 10 million unemployed, and produce viable…
SEE ALL COMMENTS WRITE A COMMENT
The very act of quantifying certain characteristics may also give a false sense of precision that leads to overweighting the things quantified. “You are what you measure,” warned Dan Ariely, an economist at Duke. Ariely thinks the hiring process should become more data-driven, but acknowledges that applying sabermetrics, or the sort of empiricism that helped turn Billy Beane’s Oakland A’s into a winner, to the recruiting process can still create distortions. Online dating sites, for instance, often overemphasize measurable and sortable attributes (like height and income) at the expense of other ineffable ones that might be more useful or relevant. “If I want to hire people for my basketball team, it’s easy to tell who’s seven feet tall and who can shoot the ball really well,” explained Alvin Roth, a Stanford economist who shared a Nobel Prize for his work on market design and recently joined Knack’s board. “The hard thing is figuring out when they’re on the court, how will the rest of the team do?”
Human beings still beat computers at detecting these sorts of soft skills, like empathy. But the people-analytics industry is hot on their tail. Right now, the gaps and glitches in these various vetting systems may encourage companies to subject their potential hires, like Daron, to even more opaque tests than a mystery-illness patient on “House.” But soon enough, perhaps, some firm might convincingly prove that a single, simple algorithm can replace the good old-fashioned grill session.
That might sound a little scary, but automating and expediting the hiring process, if done successfully, offers huge benefits. Three-quarters of the nation’s wealth is in the form of human capital; the talent and training of workers, in other words, offers far more value to the overall economy than anything else. Better allocation and investment could provide huge returns for everyone, which is particularly important at a time when 10 million Americans can’t find work and extended unemployment can deteriorate their skills. The faster we can place America’s workers in jobs they’re good at — even if they need to play a video game to get there — the better.