Falling for the Job
Earlier this year, eHarmony, the dating website, announced a curious line extension: it had begun work on a job-recruiting service to debut sometime in 2014. Initial reactions ranged from skepticism to outright sarcasm. Would HR managers hold out rose bouquets at interviews? Would "long walks on the beach" displace "efficient multitasking" as the resume pablum du jour? What did hiring have to do with dating, anyway?
Well, quite a lot, actually. To people with deep knowledge of the hiring and dating markets, it's not a stretch. After all, both involve long, costly search processes. Both involve agents with complex sets of criteria on either side. And though money can play a role, it's not the central factor in dating or in hiring. The companies applying matchmaking algorithms to job recruitment are yet another case of a disruptive technology shattering the status quo. Hiring is a expensive process, which is why it's a $400 billion global industry. In the upstarts' view, it's being approached from the wrong direction: soft skills and "cultural fit" can be better predictors of a good hire than education and experience.
Whether you're trying to find a mate or a systems engineer, the process is complicated. In economics, it's called a two-sided matching problem, and for decades, economists have considered dating and hiring so similar as to be roughly synonymous.
"When I write out an economic model," says Utku Ünver, a market design economist at Boston College, "sometimes I call [the variables] man and woman, and sometimes I call them firm and worker."
Perhaps the most salient common feature of the dating and hiring markets right now is how broken they are. Each suffers from its own grim failure rate: divorce for lovers, churn for workers. The divorce rate is estimated to be about 45%, higher if unofficial separations are included. Though the weak job market has lowered churn, as recently as 2007 about 3 million workers were voluntarily leaving their jobs each month. Today, fully 70% are said to be dissatisfied in their jobs. Both markets are wildly inefficient and plagued by discredited notions of what creates enduring matches to begin with.
Can technology create an efficient market in sectors where "soft" human factors are so much in play? In the dating world, eHarmony believes it already has. The company was founded more than a decade ago by a marriage counselor who concluded that many couples rush into matrimony over superficial matters—physical attraction and shared interests, say—while ignoring the deeper factors that create long-term stability.
Unlike other dating sites, eHarmony doesn't allow users to check out who has the cutest picture; rather, users must fill out lengthy questionnaires. Then the software yenta serves up the matches it believes will have staying power. In June, research published in the Proceedings of the National Academy of Sciences found that eHarmony-produced marriages were more resilient than those forged through other means. Some skeptics questioned the findings' significance, wondering if eHarmony simply attracts the very people who are already most committed to the idea of marital stability. But the study does pose a provocative idea. "These data suggest that the Internet may be altering the dynamics and outcomes of marriage itself," wrote John Cacioppo, a neuroscientist and the study's lead author.
The eHarmony of Jobs
In the less emotional world of hiring, eHarmony thinks its special sauce can be applied more effectively than conventional approaches. And it is not alone. Paul Basile, the CEO of recruiting service Matchpoint Careers, is among the many contenders who already pitch their startup as the "eHarmony of jobs" (a phrase with 17,900 Google hits). Basile says that just as barflies are too quick to elope with the guy or girl with the great smile and a shared affinity for Indian food, companies also focus on traits that just don't matter.
And the real shocker is that education and experience are worth surprisingly little, says Basile. Cognitive ability, personality traits, and the nebulous notion of "cultural fit" produce a better candidate. Basile tests for each but concedes that the data is still squishy on the culture question.
"Fit is not nearly as well developed as the other topics," he says.
That could be a problem for Basile, because the consensus among recruiting startups seems to be that while fit is the most ill-defined of factors, it may well be the most important in reducing churn. Amy Kristof-Brown, an expert on "person-environment fit" at the Tippie College of Business at the University of Iowa, found that an optimal fit reduces turnover and predicts whether an applicant will be a "good organizational citizen," going above and beyond the basic job requirements.
The Corporate-Culture Quest
When Brent Daily quit his job five years ago—a bad fit—he began looking into Kristof-Brown's research. He called her and learned that her work was floating around academia but that few businesses had acted on it.
"It floored me," he recalls. "It was too good to be collecting dust."
Daily cofounded a company called RoundPegg (which counts Kristof-Brown as an adviser) that is dedicated to hacking this question of what it means to fit a company's culture.
We're familiar, by now, with stories about how Google employees whisk to the cafeteria on their scooters for a glass of agua fresca. But is this corporate culture? If you like to play tennis and a company lets you wear shorts to the office, is that sufficient to determine a fit?
Daily thinks a real analysis of the DNA of a company's culture requires a deeper, wider view. RoundPegg believes culture can be measured using surveys to get at the beliefs and attitudes of individual employees. If corporations are people, as a certain presidential candidate once declared, then that's equally true of corporate culture.
"Every interaction with a colleague—from the person manning the front desk to the person sweeping floors to the CEO—all that behavior starts to paint a picture of what it's like to work there," says Daily. RoundPegg's methods also allow it to identify subcultures; if you're working in accounts receivable, the management style of the communications chief is less relevant to you.
That's not to say the scooters don't matter but merely that they're "artifacts of culture," says Daily, rather than the culture itself. "We're also a technology company," he concedes, "and we've got the requisite foosball table. But I'd say our core values are sharing information, creativity, and high performance." And he measures each of those, just as eHarmony measures the various traits it thinks matter in a romantic partnership: things like conflict-resolution style, altruism, and autonomy.
Even in this problem-within-the-problem of measuring corporate culture, RoundPegg has its share of competitors. A startup called Good.co emerges from beta soon and claims expertise in a particularly vexing segment of the labor market: millennials—those born in the '80s or '90s who are entering the workforce en masse. Churn is more of an issue among this cohort, according to HR groups. Perhaps that's because of an essential narcissism some commentators have eagerly attributed to our 20-somethings. Or perhaps it's as Good.co CEO Samar Birwadker, 33, suspects: millennials believe that work should not only offer a paycheck but also be an expression of "who they are, what they believe, and what they stand for."
Either way, if there's one thing universally agreed on, it's that millennials like to be entertained, leading Good.co to make its surveys essentially gamelike. ("Is your manager's approach like a boxer or fencer?" "If your company were a superhero, would it be called Robotica or Chaotica?")
These waters remain less charted, as getting companies to take a hard look at culture is difficult. RoundPegg's Daily, for instance, says he still wrestles with the question of whether to weight the values of senior management more heavily than those of other workers in trying to assess a company's overall culture. Questions like these may keep the workplace matchmaking sector in a kind of beta for some time, but Iowa's Kristof-Brown thinks assessment tools are already eclipsing the industry standard: "There's no reason you should still be relying on, 'Does this face-to-face interview make me think this person is a good fit?'"
Daily puts it more starkly. "I firmly believe that in 10 to 15 years we'll all look back at the early 2000s and the times before, will scratch our heads and think, 'What were we doing ...?' It's the equivalent of using leeches to cure disease."
As in any emerging sector, the struggle to find the right formula causes flameouts. One company, Path.to, drummed up more than a million bucks in funding last year only to shut down this summer. "Unfortunately, we have learned just how hard it is to innovate in the hiring space," the company wrote in a July 16 blog post.
One not insignificant issue is how these outfits will make money. Many are relying on variants of the traditional model—charging businesses for their services or specifically when a hire is made. Others are trying different pricing: RoundPegg, for instance, offers companies a culture analysis for a one-time fee—from $5,000 for a small organization up to hundreds of thousands for huge ones. Once a matchmaking firm reaches scale, something a giant like eHarmony could achieve quickly, the real cash would come from monthly subscriptions. LinkedIn, which makes most of its money as a recruiting tool, already does something similar.
The more-cautious innovators in the hiring market believe in the long-term matchmaking promise of big data but are reluctant to wait for the computer oomph to become available. GroupTalent is a marketplace for software-engineering jobs and relies on lightweight interviews to assist its matchmaking. But cofounder Manuel Medina says he doesn't think the science is there yet to create lasting matches based on personality tests and surveys alone.
As he grew his business, one of its incidental features became a hit—brokering "try before you buy" deals, in which workers and companies test compatibility on-site for a few weeks before committing to a long-term relationship. Clients kept telling Medina that the arrangement reminded them of something, and eventually Medina had to agree. "Get paid to date companies," his website now proclaims.