Like dating, a human endeavor frought with uncertainty, reward, and high turnover, hiring has a habit of taking all the theories and recommendations you have about a candidate (or, from the other side, the company) and emptying them unceremoniously into the nearest garbage can. At a fundamental level, when two people sit down to an interview, it's two humans trying to figure out if they like each other.
And, according to Northwestern researcher Lauren Rivera, what most people are looking for is "me." In the companies she studied, interviewers who lacked systematic measures of what their company was looking for tended to fall back on themselves and defining merit in "their own image," meaning that the most qualified interviewees were those who best resembled their interviewers.
Because of this pattern, the consultancies, banks, and law firms that Rivera studied tended to "replicate themselves," hiring only people who had the same hobbies, styles of self-presentation, and educational and geographic backgrounds. Even if there was nontrivial demographic diversity across race or gender, there remained a more fundamental homogeniety of culture or personality.
Which is a bad thing for innovation. Research shows that if you want to create new ideas, you need to include the outgroup.
What Rivera's studies evidence is that the murkier aspects of hiring—like personality fit—have been left to hunches, heuristics, and their inevitable human biases, a multi-faceted blindspot that a crop of companies seek to cure.
You may have heard of eHarmony. Launched back in 2000, the Santa Monica company known for making long-lasting couples—to the tune of 100,000 marriages a year—is now taking first steps into the recruitment racket. VP of customer experience Grant Langston tells Fast Company that they'll probably bring a product to market next year, though at present the effort is being developed by four full-time employees and a gaggle of consultants.
Langston says the company will try to slow down the rate of job-hopping, similar to what they did with divorce. He chalks the turnover rate up to superficial assessments of job fitness for both employee and employer, leading to fundamental disconnects and leaving employees feeling underappreciated.
"The jobseeker, more than anything, craves the knowledge that they end up in a place that values who they are," he says, comparing that with the sense of my-partner-understands-me that the romantic matching has enabled. "If we can put that same experience in a job situation, we’ll have a lot more productivity and happier people."
When Sean Glass was getting his masters in Applied Positive Psychology at Penn, he was struck by how good scientists had become at grabbing the data underlying what's good about them—emotional intelligence, psychological strengths, and the like—and that no one had built software built around that data. His company, EmployInsight, is aiming to change that by allowing you to profile your work along a psychological dimension, both by assessment engine and matching algorithms, and in doing so helping to cure that psychological blindspot (for a consumer-facing idea of how this works, check out StrengthsInsight). "Just as you would specify the skills for every job you hire for," Glass writes to Fast Company via email, "you should also work to specify the psychological resources used by that job."
Glass makes the following case: While you may need the same technical skillset for every engineer on your team, you'd want to have a diversity of psychological resources—one should be more creative, another diligent, another a quick learner. "It was hard to specify jobs this way before," he says. "We've made it much easier."
Before you can get to personality, you need to start with resumes, says Bright, who combed through 2 million of them (and 15 million job descriptions) to sculpt their algorithmic Bright Score, which matches skillset and experience to position.
"If you’re a call center and you need to hire for the next two months, its really hard to date 2,000 people at the same time," says CEO Steve Goodman, "so the first order of business, regardless of cultural fit, is getting the short list of people. Once you have that, then you can date."
The Brightscore doesn’t take personality questions into account, says chief scientist David Hardtke, who says that personality characteristics can reduce competitiveness of applicants. He mentions a Fortune 500 company that Bright consulted for that wanted "go-getters" to be their salespeople—so they only hired fraternity and sorority presidents as their salespeople, eliminating candidates and reducing diversity.
Still, Hardtke says, the most pressing of questions—who will be the best performers in a given role—is hard to answer (and to collect information for). Hardtke says that it is "very difficult" to gather sufficient systematic data to predict person-job fit; person-organization is much easier. As Bright grows, they have the goal of collecting enough data from enough corporations to improve person-job fit—though it's not there yet.
Founded in 2011, jobFig uses the five-factor model of personality to measure compabililty between pre-existing teams and potential hires. Using jobFig, you can take the 100 people who applied to your last posting and reduce the number to those that will get along with the team—and then interview them.
Specializing in interpersonal dynamics, the Mountain View startup tracks leadership and work styles. Using their personality mapping tech, the startup can measure where there's a leadership vacuum or a lack of implementers. If you've already got a leader, jobFig will filter out the leader-types so the two don't fight like betta fish. Using their personality data crunching, the qualitative ambiguities of team cohesion grow more quantitative and actionable—similarly to the other startups in the prehire assessment space.
To cofounder Ravi Mikkelsen, it's a positive: "Everyone in this space is competitive in that we’re all prehire assessments," he says, "but we’re also quite complementary in that we all do something slightly different."
Still, the challenge is in the data-collecting itself.
"In order to do better hiring, you need to have a really solid, systematic, and scientifically valid understanding of what it takes to be successful," says Rivera, the Northwestern assistant professor. "So until you have that, you can match people to the cows come home, and I’m not confident it would actually give you better or worse hires."