Hiring decisions have always been limited to a few imperfect factors, including what appears on a resume and what impression a candidate gives off in an informal interview.
That's all changing.
As Dan Shapero, LinkedIn's VP of Talent Solutions and Insights puts it, "Recruiting has always been an art, but it’s becoming a science."
Using new tracking and analytic tools, researchers have learned to value things like adaptability, social and emotional intelligence, resilience, and friendliness, as well as raw intelligence.
Companies are scouring the Internet for data about potential employee's professional lives, applying the big data technology to years of employee surveys and tests, and even picking up new data from specially designed games.
A big data approach to hiring and management has been embraced by companies like Google, which has an entire department devoted to "people analytics," according to the New York Times.
In the near future, it will change hiring practices across the board. Just look at what similar insights did to professional baseball, as described in Michael Lewis' "Moneyball."
The death of the résumé
The Moneyball revolution in baseball led to lots of old data, tools, and assumptions getting thrown out. It was found, for example, that runs batted in (RBI), batting average, and how many wins a pitcher has are not very useful statistics. The same is true for the things that used to matter most to recruiters and managers.
Take the résumé, long the be all and end all for for employability.
They are relics of the dark ages of recruiting, where big recruiting firms had the only real databases of potential hires, according to Peter Kazanjy, the CEO of TalentBin, a startup which scours the web for potential technology hires.
"There are a couple of problems with résumés that make them not great," Kazanjy argues. "One, how the hell do you get access to them? And two, they're not necessarily as data rich as you might expect."
Since most people, especially the most in-demand, aren't actively searching, a big hunk of the recruiting process happens before the résumé enters the equation.
"Résumés are actually curious constructs now, because for the most part, work and our work product is fundamentally digital," Kazanjy says. "Sometimes you don't even need [résumés]. The reality of what somebody is and what they do ... is already resident on their hard drive or their Evernote or their box.net account or their Dropbox cloud."
Much of that information's private. Most people have enough of an online professional presence, however, that it's possible to build a surprisingly complete picture.
Companies are using big data sources like TalentBin and tools like LinkedIn Recruiter to go out and find passives, learn about them, and use contact points like a shared connection on LinkedIn, a colleague who went to school with a candidate, or data on their interests to make a more engaging approach.
In fact, the data and services LinkedIn sells to recruiters is one of its biggest businesses, accounting for 57% of its revenue and growing 80% year-over-year.
LinkedIn solved part of the supply and demand problem by making the résumé a public, rather than private document. Recently it took another step in going beyond the résumé by letting people post pictures, slideshows and other content on their professional profiles.
Thinking beyond college
Recruiters used to treat college prestige and grades as a critical factor. According to the data, however that's a mistake.
Guy Halfteck is the founder and CEO of Knack, a Silicon Valley startup which couples games with advanced data analysis to figure out what sort of characteristics make for successful and innovative employees. Clients include Shell, which uses the technology to identify potential innovators, and NYU's department of orthopedics.
"If companies are trying to quantify success based on which school you went to and what's the ranking of the school and your GPA, etc. those things are not very powerful," Halfteck says. "You might get into a school because your father got into the school. Those are not indicative and insightful about who you are as a person and your potential."
Far better to do an analysis based on observing actual behaviors and performance in real time via one of their games, he says. Fifteen minutes of play is enough to create a megabyte of data.
Good college grades indicate a certain level of intelligence and diligence, but maybe it's the wrong kind. The highest-value jobs don't come with a ton of instructions. They require adaptation and learning on the job.
"Companies want someone who thrives on challenge [and is] willing to learn something new," said Josh Bersin, the principal and founder of Bersin by Deloitte, a data-centric HR consultancy. "[They want] a seeker of information, willing to adapt. If you're the type of person that wants to be told what to do, you might be a straight A student. In fact you might even be a better student than the other type of person."
Prasad Setty, Google's vice president for people analytics, recently told The New York Times that they've found that SAT scores and college GPA don't lead to success at the company, and "are no longer used as important hiring criteria."
A name brand school matters in as far as it pre-screens for smart, high-achieving students, but it doesn't mean nearly as much as people think. Put to the test, traditional criterion don't hold up, Bersin says.
"We have a client — an insurance company. They're a Northeast-based sort of traditional Ivy League-led company. They have this culture of only hiring people [who] had MBAs from top schools and so forth," Bersin says. "They looked at the performance of best sales people and they found it had nothing to do with where they went to school, and nothing to do with their grades."
Even their references didn't mean all that much. The ability to perform well in an undefined job means much more.
Replacing the interview
Interviews, which play a role in pretty much every hiring decision, are an exceptionally poor source of data.
It's an information problem, Knack CEO Guy Halfteck says. "How unique can these questions be? You tend to get more of the same."
"The ability to attract insights — and predictive insights — and make an intelligent decision based on that info is very poor," Halfteck argues. "People fake it, they put on a show, we all know that interviews are not predictive."
Not only are interviews bad sources of data, people aren't all that great at them. "Most people don't know how to do good interviews," Josh Bersin says. "When most people do an interview they say 'oh I really like that guy,' well you might like him but that doesn't mean he's right for the job."
Interviews are fraught with irrationality. We select people who are good looking, who look or think like us, or that we kind of like.
Computers don't have that problem.
Better data
The other part of the "Moneyball" equation is replacing what doesn't work with something that does.
To start, there's the vast selection of professional data available about people far beyond their résumés.
A software engineer might not mention on LinkedIn that they work with iOS. But that doesn't mean you can't find them, says TalentBin's Peter Kazanjy. It's about looking at the massive amount of data people put in the Internet, and taking what he calls the "Google approach" and using technology to collect and make sense of it.
"Humans engage in professional activity all across the web," Kazanjy says. "Even if [an engineer] never put it on a LinkedIn profile you can know that he's been tweeting about iOS and Xcode and Objective C, that he's a member of meetups on meetup.com or has answered questions on Stack Overflow, or he's following a bunch of repositories on GitHub or he's participating in Apple engineering support email chat lists about Xcode."
"That's an iOS engineer, obviously," Kazanjy says, "but it's obvious only when you've gone and looked at the implicit activity and said 'ok cool, walks like a duck quacks like a duck, must be a duck'"
That lets you get beyond what people list as their skills and interests and find the people who are passionate and engaged with the communities around those skills.
Gild, another startup, looks specifically at the quality of the code engineers put up on GitHub, a popular hosting service for software development projects, to find diamonds in the rough.
It's even possible to get incredibly useful data out of the way people play games, which may involve intense engagement, reaction, and even reading other people.
Knack has built a variety of games that help them find the behaviors and personality traits that boost productivity and performance, to "quantify success," in the words of CEO Guy Halfteck.
"Cognitive ability is a small fraction of what we measure," Halfteck says. "We measure everything from creative abilities to emotional and social intelligence, to how you think and make decisions ... how you interact with emotions, understand emotions, how you learn new information, how curious you are about the world."
Halfteck calls it "behavioral big data." They'll work with a company and build skill profiles by having a variety of employees play their games. That way, they can find what makes for a successful innovator, manager, or engineer at any given place. A computer can get useful data from every instant of a user's engagement with a game, is better at finding patterns, and is more objective than any human alternative.
The power of big data technology and advances in business psychology means that data already stored and information already collected through surveys and assessments are more useful now than ever. Kenexa, which surveys and assesses some 40 million people a year, was recently bought by IBM for more than $1 billion to boost its capabilities in this growing field.
"What we're bringing to the table is what we call the human insight analytics," Kenexa CMO Tim Geisert says. "So what drives engagement from an organizational standpoint? What are the key data points that make people good at what they do in their jobs?"
"What's supercharged that is the technology that can gather that data," Geisert added. "The platforms which we run that data on and using it for analytics, insights, and predictability."
What employers are looking for today
Adaptability
"Every HR person I talk to says that your passion and drive overcomes educational background and ability except for one thing," Bersin told Business Insider.
The one innate quality that is absolutely necessary for every job is "learning agility." That's the ability to pick things up quickly, to learn on the job, and to take initiative.
Over and over again, the people who perform best are the ones who don't need to be told what to do, the ones that love challenges, seek information on their own, and quickly adapt.
People who follow instructions are mostly substitutable. Those who can be thrown into a new situation and thrive are truly valuable.
Resilience
Stereotypes about certain jobs are often proved false when you look at the data.
"Most people believe that a good salesperson has to be an outgoing personality, has to be really friendly and good with relationships," Kenexa CMO Tim Geisert says. "Part of that's true, but what we've found in our data is that there's actually a trait that's hidden that predicts more success than any of those other more overt traits and that's called emotional courage."
It's being resilient, being able to hear "no" again and again and keep going.
Social and emotional intelligence
Raw analytic power isn't everything. And the need to work well with others isn't just needed for those in customer service.
Knack has found that one of the major things that correlates with success across just about any job is social intelligence.
"One way to think about it is that everything we do, and try to achieve inside organizations, requires interactions with others," CEO Guy Halfteck says. "Whether you're an innovator, a physician, a teacher, a retailer, or a salesperson, your social abilities, being able to intelligently manage the social landscape, intelligently respond to other people, read the social situation and reason with social savviness — this turns out to differentiate between people who do better and people who don't do as well."
If you come up with an innovative idea, it likely won't go anywhere if you can't convince anybody. It's not just about creativity.
Knack measures aspects of social intelligence, with, among other things, a game called "Wasabi Waiter" where job applicants play as a server and are measured on how well they read social and emotional signals.
A diverse background
One of the most difficult jobs to fill for oil companies are production roles — the ones where people spend years out in the field, often living in the desert.
When one oil company looked at the data for who stuck these jobs out and succeeded, they found surprising results, Josh Bersin says.
The company started out the traditional way, looking for people with petroleum degrees, good academic credentials, and so on. But when they looked at the data, that didn't predict success.
"The head of recruiting found out that the people who were surviving in these jobs — and these are jobs where the turnover rate was very very high — were mostly people who had come from families that had multicultural parents, parents that had international experience," Bersin says. "They grew up in a climate with lots of different types of people around, which wouldn't necessarily be true depending on your college."
Another interesting find? The most successful people had played sports in college too.
Friendliness
One theater chain that Josh Bersin worked with found a surprising amount of variation in concession sales from theater to theater. When they tried to track down the cause, they found that employees were happier at the high-grossing theaters, and that customers were more satisfied.
Naturally, they tried to boost that in all of their stores by training everybody in better service, but six to nine months of training didn't produce results. It wasn't the training, it was the people, the head of HR discovered.
"There are people who are wired to be in customers service; they like it, they like people. There are others who aren't," Bersin says.
The company's hiring practices were based on the usual: grades and credentials and degrees. They tweaked their screening criteria and pre-hire assessment to focus on whether people had happy personalities, enjoyed being around people, and liked serving others. The return on that investment was in the millions of dollars.
Raw processing power
At the end of the day, how sharp you are still matters. And so does how fundamentally diligent and careful you are — whether you always want to do a thing well.
Through the many games Guy Halfteck and his team have put together, those have been two of the constants.
"We know certain things about the world. We know that across the entire set of jobs that people can engage in, it's general intelligence that's your processing power and conscientiousness. That you do the right thing and care and plan ahead — those things — and to do the right thing," Halfteck says. "These are independent of motivation. Those two aspects of a person are very predictive of performance across any type of job."
Increasingly, a high score in baseline attributes like these will mean more and more as companies look for raw potential.
A professional presence
With more data and more tools available, more employers are looking for what's known as "passive" candidates, people that aren't actively job hunting, but might be perfect fits. There's an increasing possibility of getting approached for a better job, even if you aren't looking.
But they have a harder time finding you if your professional presence online is sparse. Recruiters are increasingly interested in finding the best candidate based on data, skills, and qualification . That means the bar is higher, says LinkedIn's Dan Shapero.
"Since recruiters can now proactively find relevant candidates, they have more choices and can be more selective," Shapero says. "Rather than sifting through hundreds of irrelevant resumes, recruiter can selectively choose who to target and find best candidate for a job."
"As a result, recruiters now expect candidates to have a professional presence online that showcases their professional brand," Shapero adds. "It’s in candidates’ best interest to showcase their professional brands online because the majority of recruiters no longer post requisitions on job boards and pray that relevant candidates will apply. They’re using recruiting products to proactively search professional sites, like LinkedIn, to find the best and brightest."
In addition to third parties keeping an eye out, according to Shapero, companies like Cisco are using LinkedIn's platform for internal mobility, to identify high performers, identify what they're interested in, and helping advance their careers, which is helpful in retaining top talent.
