Should Big Data Change The Way You Hire, Fire And Promote?

by Grant Gordon

The lead article in this month’s issue of The Atlantic, “They’re Watching You at Work,” focuses on how Big Data is helping transform HR processes in all industries. In the course of the story, writer Don Peck explores how the kind of analytics that powered the Oakland A’s to success on a shoestring (and made General Manager Billy Beane in the movie Moneyball) is changing the way companies evaluate candidates.

There’s no doubt that a data-driven approach is taking hold, particularly for larger companies with the deep pockets to buy the software and hire the people to interpret the numbers it spits out like fastballs (think more like the Yankees than the A’s). Specifically, Peck mentions Google, HP, Intel, General Motors, and Procter & Gamble.

Certainly such companie are making good use of the analytics their investments provide, but very few organizations have the money or the staff expertise to go as far down the Big Data road as those mentioned in the article. That’s certainly a limiting factor for startups and midsize companies. But I think an even bigger challenge with taking a number-first approach to hiring, firing and promoting is the notion that the “elimination of the human” is the desirable next step. These are not the writer’s words, by the way, but those of one of the readers who commented on the story.

After years spent working face-to-face with IT organizations and the technical talent who fill their vacancies, I can assure you that there are many things a data-driven method won’t tell you. It’s only by sitting down with company leaders that you come to understand their culture and, conversely, by meeting with IT professionals that you can tell who would thrive and who would wilt in such an enviroment.

Once you’ve identified a candidate who’s a good match with a client, the organization can’t just take your word for t or look at aptitude tests. They need to meet the person and find out who tey are, what they can contribute and, hopefully, why they’re going to be the right long-term fit. The writer, describing a case study from a large tech company, shares a troubling quote from that company’s HR manager: “We’re getting to the point where some of our hiring managers don’t even want to interview anymore – they just want to hire the people with the highest scores”.

If we ever reach the point in the HR industry when this view becomes the norm, something valuable will be washed away forever by streams of data. As Peck correctly observes, Google is arguably the king of gathering data and applying it to hiring practices, and yet even its hiring process is still “deeply human”. While computers can certainly add objectivity to talent evaluation, they cannot take the place of an experiences HR team and the people-first technical staffing organizations they often collaborate with to identify and place top talent.

This is not to say that objective, data-centric tools should be completely ignored – rather, they can be incorporated into a people-first approach to help technical staffers and their clients. For instance:

  • Find local candidates who are the best fit for o-site positions for which relocation reimbursement is not an option
  • Sift through a nationwide applicant pool to find those whose experience, interest and personality makes them right for remote/telecommuting roles (such as how Google is hiring a lot of people to code from the Midwest).
  • Discover which candidates are active and respected in their specialized area, such as .net development, by scoring them on how many posts they make on expert forums and how many people are using their code.
  • Evaluate and aggregate activity on social media sites such as Stack Overflow to determinte who’s at the top of their game.

By using methods like those above, companies can shorten the time-consuming elements of resume review and applicant screening in order to deliver small, selective, targeted final pool that the HR team can then conduct phone and in-person interviews with. It’s the best for both worlds and, in my experience, one that yields a much better end result. Because whatever benefits raw data provides, your gut instinct from a 15-minutes meeting with a candidate can tell you more than years of performance analytics ever could.

Extracted from: