You see articles like this all the time, where some academics run a relatively-complicated analysis of hiring and recruiting practices through the lens of diversity, and then propose at the end that HR and TA should be more “data-driven” and figure out where their blind spots are with intel.
All well and good, but can HR do that?
I went to a grad program (stupidly, as I’m still in debt) from 2012–2014, and a lot of people who emerged from that took HR roles.
Whenever we had a big data project, like around compensation modeling or High Performance Employee indicators, everyone freaked out. One kid, whose name I think was James (?), was good at the stuff and ended up getting a sick job with United Healthcare doing data analysis within the HR silo. I have no idea where he is now; haven’t thought of him in 6.5 years.
The people that enter HR tend — not always, but tend — to not be very data-savvy or analytical. If you were a true data ninja, why would you ever come in through the HR silo? You can chase more money coming in related to product, operations, sales, etc.
Usually HR is women, often women pre-children, and periodically it’s people who define themselves as “a people person” and then perp walk you to the curb with a cardboard box of your crap. So that’s pretty fun, and well-branded.
You don’t see a lot of data ninjas in that world. Some, but not a ton.
Then in standard recruiting or TA, you have a lot of phone warriors, smile and dial people, biz card chuckers, glad-handers, and some more cute blond girls around age 26.
Again, not a lot of data ninjas. Some. But not a lot.
So who exactly is supposed to be running all this diversity data for hiring silos?
Execs? They broadly don’t care. Their data efforts are aimed at products and financials.
A data scientist in-house? Maybe, but his/her attention and priorities will get pulled off HR/hiring about 1,000 times every three months if something related to revenue crops up.
We keep saying “data-driven” will solve our hiring and HR issues, but there are tons of problems with metrics from the HR silo, and there have been for years. We still haven’t really scaled People Analytics, honestly. And just this morning in a newsletter, I saw this:
The percentage of HR managers who said their company uses predictive analytics algorithms in the course of hiring rose from 10 percent in 2016 to 39 percent in 2020. AI-based hiring is faster than human-based hiring, but researchers have found that these systems not only reflect the racial and gender biases of those who train them, but also entrench it and hide it behind math. As a result, companies now have to bring in algorithmic analysts to ensure that their tests don’t incidentally discriminate and violate the law.
So what’s the answer in the HR silo, commonly?
Automate it. Use machines. Get these pesky HR humans out the door. (You know that’s what most execs want: “Defund HR.”)
Again, though, who is running these data sets, scrubbing the data, organizing it, and getting on the calendar of decision-makers?
If that’s happening, awesome.
I doubt it is at scale, honestly.
So data cannot be the path through on these hiring issues, because while data should ideally knock down subjectivity in a 12-round fight, that’s not always the reality.
With execs looking to fully automate HR by 2030 if not sooner, and diversity initiatives basically running in circles, have we actually proven yet — even remotely — that data can fix hiring issues?
I don’t see it. Do you?