"I'm sorry, Dave. I'm afraid I can't do that." - HAL 9000, 2001: A Space Odyssey
Remember the Wall Street Journal article I told you I was interviewed for? Yeah, well turns out I wasn't quoted, probably because I was a big wet blanket on the topic. But that doesn't matter--what does matter is continuing the conversation that has been sparked by the piece.
The WSJ article talks about using analytics and computers to hone in on what makes for good hiring decisions. In particular, using computers to develop models for the best hiring algorithm. Nothing wrong with that (well, conceptually)--I/O psychologists have been doing that for a long time. I just thought it was sorta old news. But other writers have taken interesting approaches, talking about the implications of "big data" and, well, "big computers", on hiring.
Charles Handler makes some great points in his article on this (including the importance of human judgment and process), as does Joe Murphy in his (e.g., a focus on validation), but I'm going to take a slightly different approach.
Good hiring doesn't hinge on technology. Or validity (bear with me). Or job analysis.
In the practical world, at least in mid- and large-size organizations, good hiring is about the quality of relationship between the HR consultant/recruiter and the hiring supervisor. Because without it, you're guaranteed to introduce bias and lower your success rate.
Why? Because most supervisors aren't experts in hiring. They are, hopefully, experts in their business. But left to their own devices, they WILL ask stupid questions like what animal do you wish you could be. They will make assumptions about things like short job tenure (bad) and attending the same alma mater (good). And they will hamstring the organization from being as effective as it can be as a result of middling talent decisions. But ya know what? I don't blame them--I don't expect them to be good at hiring. That's where HR comes in. HR consultants, like good mechanics, are tasked with diagnosing the situation and coming up with effective, efficient solutions.
The power--and if necessary, blame--lies with the HR function. It controls the relationship and should be held accountable for it. Not sure how well an organization is doing? Here are a few ways of diagnosing how well this relationship is working:
1) How well do the hiring consultants know the business they support? (Yes, this comes first)
2) How well do the consultants know hiring research and best practices?
3) How much are the consultants trusted by their customers?
4) How often do supervisors contact HR about new issues, simply to get their advice?
5) How much unsolicited positive feedback is given to the consultants?
None of this is new. So why am I talking about it? Because it needs to be emphasized. Again and again and again. I've discovered that every time something "new" is discovered in hiring technology, I have to remind folks to go back to the basics.
Technology, or more accurately, good analysis, is essential to ensuring the best ROI on hiring. And as much as I like technology, as much as I have wished sometimes that certain supervisors weren't involved in hiring, they are, and that's the world we live in. And that's not going to change any time soon. Why? Primarily control and arrogance.
First, control. In some organizations, the culture is such that big data may be given priority status over human decision making. But I'm guessing in most it won't (particularly those with merit systems). Because supervisors will not relinquish control over hiring--at least the final decision, and not without a lot of resistance.
Go ahead, take this exercise: think about hiring someone to report directly to you on a daily basis. If I told you I had the perfect algorithm that predicts performance, that you don't even need to meet the person before they start...would you agree to hire them, sight unseen? If you would, kudos: you're unlike 99% of the people I've asked that question.
On to the second part: arrogance. Even after years and years of research with large data sets indicating otherwise, most hiring supervisors believe they know how to judge people and make good hiring decisions. They don't. One thing the focus on big data and the I/O research have in common is a recognition that the more "objective", consistent, and standardized (i.e., less influenced by human biases and decision limitations) you make the decisions, the better they tend to be.
Put these two factors together and you get the status quo, at least for now. Rapidly evolving organizations are more likely to adapt computer-assisted decision making. But in the end, we're unlikely to remove human decision-makers from controlling final hiring decisions in many organizations. For now.
But perhaps that's not a bad thing. As bad as our biases can be, the errors that entire computer systems can cause are much more serious. If only I could think of an overblown example...
"The system goes on-line August 4th, 1997. Human decisions are removed
from strategic defense. Skynet begins to learn at a geometric rate. It
becomes self-aware at 2:14 a.m. Eastern time, August 29th." - Terminator 2, Judgment Day