July 29, 2014

“Interviewing Should Be a Privilege”: Can Big Data Discover Your Company’s Best Interviewers?

When we think about the hiring process, we’re often concerned with the candidates first and foremost: who among them is the best? But what if we started thinking about our hiring managers and recruiters, our HR departments and our interviewers — what if we started thinking about the people who were reviewing candidates to fill the open position: who among them is the best at hiring?

Mark Newman, founder and CEO of digital interviewing platform HireVue, believes we need to start asking that crucial question. According to Newman, HireVue found that the “batting averages” of talent identifiers inside of any given organization can range drastically, from 22 percent to 85 percent. “You can imagine the size of a company that’s hiring ten thousand people or a thousand people, or a hundred people,” Newman says. “What if 60 percent of the hiring decisions that had been made were made by people with less than 40 percent hiring accuracy? That’s screwed up! That’s not good for candidates, that’s not good for the company, that’s not good for the managers.”

Part of the problem, says Newman, is that data regarding who is and who isn’t a good judge of talent has largely been anecdotal. “The worst thing in the world is when you have that manager who’s like, ‘Oh, I’ll tell you how to hire people!’” Newman says. “And you’re like, ‘Hey, whatever that guy votes, we’re gonna do the opposite. If that guy says don’t hire this person, we’re gonna hire them.’”

In fact, much of the crucial data in the hiring process has been anecdotal, offline, and “in your head,” according to Newman. Think, for example, about the traditional interview process: “You interview me, you take some notes, and you have some kind of context in your mind around how you’re thinking about it,” Neman says. “As you go on over time, 30, 60, 90, 180 days later, a year later, multiple years later, you never actually can go back to it. Your own personal hard drive is gone, in the sense of what’s in your brain. You’ve figured out a lot of other things, you’ve seen stuff go on, and data that should exist in an applicant tracking system or talent system or whatever doesn’t actually exist. It’s very sanitized.”

To recap: companies have no real data to track who their best hirers are, and hirers have no real data regarding who the best hires are. “This data has never really existed and been in the hands of individuals to actually do something with it,” Newman says.

The solution to this predicament, according to Newman, is big data. This is why HireVue has announced HireVue Insights, which bills itself as “the first predictive candidate and interviewer recommendation engine.”

From Video Interviews to Big Data

“I think people need to be looking at [big data] as actually tying together disparate data sources that previously were totally disconnected and really thought to be kind of random or not related, and mining those to see if there actually is something interesting that comes out of it,” Newman says.

With this conception of big data in mind, it’s a short walk from HireVue’s digital interviewing platform to HireVue Insights, the big data-crunching machine the company recently released. HireVue is a sort of DVR for interviewing, says Newman. The platform records candidates’ answers to everything — via video, text, etc. — effectively allowing interviewers to view interviews on demand. Powering thousands of interviews for a wide variety of companies allows HireVue to collect large amounts of data about candidates, including feedback on candidate experience. HireVue can also track who did the hiring at each company, as well as the performances of the candidates who were hired.

Explaining what HireVue tracks, Newman elaborates: “What actually happened? Did you hire them, did you not hire them? Did they come onboard, did they turn over, did they not turn over? Did they become top performers?” And we have all sorts of other things wrapped around it. We can automatically measure motivation, engagement, stress, personality style, and all these various things around these interviews.”

Tracking all of these things has allowed HireVue to create “a really unique data set,” says Newman.

Thinking about it in terms of Newman’s definition of big data, HireVue has built a huge pool of “disparate data sources,” which it now seeks to tie together through HireVue Insights, “this engine behind it all that runs this massive analysis against [the data],” according to Newman.

Big Data and Your Top Identifiers of Talent

Part of what HireVue insights aims to do is predict which candidates will be top performers at a given organization based an impressive range of interactional, behavioral and performance attributes. “There are 15,000 attributes inside of an interviewee, inside of a single person — like the essence of who somebody is, the good stuff — not just kind of what they wrote up on a resumé,” explains Newman. “We take the outcomes of actual top performers inside of an organization, and we run this massive analysis around those 15,000 attributes. What are the ones that actually matter? Is it 7,000 of them, or is it 22 of them?”

By figuring out which attributes your company’s top performers demonstrate, HireVue insight hopes to match candidates’ attributes against the ideal template and predict who will go on to succeed at your organization.

The idea of using big data to predict successful employees is nothing new. Not too long ago, I spoke to Amar Dhaliwal of Saba about the same thing. What is new about HireVue Insights, however, is its ability to determine who your hirers are — that is, who has the best hiring accuracy rate at your company? Who consistently hires the best performers on staff? “We’re going to figure out how those top identifiers make their decisions, and replicate that at mass scale, to recommend candidates,” Newman says.

Smart hiring is about more than just picking the best candidates — it’s about making sure you know how to pick them in the first place. “Interviewing should be a privilege,” Newman says. “The guys who make good hires, they should be exalted at the company.”

But HireVue Insights isn’t only good for companies, according to Newman. He also believes it poses benefits to candidates as well. Not only does it enable candidates to tell their stories, he says, but it also also allows companies to say, “you’re only going to be evaluated by top identifiers of talent. There’s no more of these people who just don’t know how to identify talent doing interviews.”

In other words, HireVue Insights seeks to make it so that incompetent hirers will no longer prevent high-performing candidates from joining the companies where they’ll soar. Newman invokes the adage that “a weed is just a flower growing in the wrong place.” “The same goes for talent,” he says. “You could be a salesperson at one company and not be doing very well, but you could be a salesperson at another company and just be knocking it out of the park.”

With HireVue Insights, Newman hopes that his company can match the right employees with the right employers. No weeds, only flowers.

Read more in Hiring Process

Matthew Kosinski is the managing editor of Recruiter.com.