In case you hadn’t heard, it turns out that recruiting is dead (or dying, or going extinct, or some similarly dire prediction) – or at least that’s many of the recruiting pundits and prognosticators out there seem to believe.
Of course, this industry has always looked at the world with a Chicken Little sort of mentality, perpetually holding onto the belief that the sky must be falling.
Look no further than your social feed, and surely you’ve seen at least a few examples of what’s become an entire genre of predicting the demise of traditional jobs, the obsolescence of sourcing and recruiting, and, of course, the rise of the robots poised inevitably (and invariably) to replace recruiters and hiring managers with algorithms and AI and automation, and so forth.
You know the drill by now.
Because while this sort of doomsday scenario inevitably drives eyeballs, clickthroughs and conversions, even the most confrontational content around this perpetual talent trending topic is starting to feel a bit tired, lazy and hackneyed. After all, we’ve been predicting the demise of recruiting for as long as I can pretty much remember, and somehow, we’re not only still here…
We’re kicking some serious hiring ass.
How To Tell A Story, And Others.
No matter what the “thought leaders” and “influencers” might think, the evidence points pretty strongly to the fact that just like that time we all thought we needed to ensure all the COBOL dependent machines out there didn’t cause the world to blow up on the stroke of midnight on January 1, 2000, we’re going to avert this preposterous prediction of apocalyptic proportions. Sometimes, it feels like Y2K is happening today – at least, if you’re in the business of hiring.
In full disclosure, that’s precisely what I do in my day job, where I lead technical recruiting for one of the world’s biggest consumer market research and analytics companies, where anything that can’t be measured in petabytes isn’t worth measuring, where we scoff when we hear “big data” because, frankly, we’re at the forefront of discovering how big, exactly, data can be. It’s our business to break down the massive amount of our data being fed into our machines every month, crunch the numbers, and then make predictions based on this analysis.
Having spent the better part of a decade of my career here, I’ve had the chance to see firsthand how our engineers work and what they do with all that data. As a recruiter, it’s my job to know the business, and since this is the business we’re in, I’d like to think that this experience has led to my knowing at least a few things about what the hell machine learning actually involves and how it works – and while I’m by no means an expert, I do know enough to know that we wouldn’t be able to do what we do without the technology we have today.
Technology is the foundation for everything we do with data.
But I’m also well aware that managing all of those petabytes and actually achieving actionable insights and accurate analysis takes more than machines. It takes people. I know, that’s getting to be a bit cliche by now, but it’s true. Without people ensuring that the machines are online and operational, without people programming and monitoring the manifold algorithms to eliminate any biased, bogus or misleading data that might influence our analytics, and without people able to effectively transform this data into compelling use cases and business strategies, our technology (and the analytics it drives) would be more or less worthless.
Data without a story, after all, is just a bunch of nonsensical numbers and meaningless metrics.2
The Gilded Age.
Recruiting, believe it or not, is predicated on pretty much the same exact premise. We’re in a function that’s historically proved willing to adopt new technologies to improve our processes and augment our capabilities, assuming that such technology is both accessible and affordable (kind of like how LinkedIn used to be, back in the day). Historically, we’ve had no problems asking algorithms to inform where we post our jobs, how candidates rank for those jobs, and which ones are worth burning through an InMail credit to send yet another generic template with an impersonal message.
And while artificial intelligence and machine learning can, of course, assess how long someone’s relative tenure has been at their respective job, and may use that information in aggregate to come to specific (and specious) conclusions around their likelihood to move based on that longevity.
For example, some of the AI/ML instances out there are likely to tell you that someone who’s 2 years, 7 months into a job has a higher probability of responding of changing roles than someone who’s only been there for 2 years and 6 months, but less likely than someone who’s been there for three and a half years and works at a startup, who’s probably waiting around for their shares to fully vest.
Such information, of course, can be useful, but it’s really only half the story – and I probably don’t need to buy a machine capable of beating Ken Jennings in Jeopardy! to figure out. Because it’s elementary, my dear Watson – no machine in the world can ever assess the personal stuff that informs and influences our professional lives.
Things like outside interests, the inflections in their voices, whether or not they actually took the time to research your company or are just reading some copy off of your website – these are things that no machine in the world is capable of. No artificial intelligence out there is smart enough to look someone in the eye and know whether or not they’re full of it.
Hiring decisions are based as much on intuition as information, as much on gut feelings as analytical insight, and these are things no machine can ever learn (even if they’re one day able to approximate it). Data might be a science, but recruiting, largely, is more art than algorithm, at least if you’re doing it right.
But although we’re beginning to move from creating overly complex Boolean strings with multiple modifiers and operators to having a platform build them for us with the push of a button (with at least a 50% sourcing success rate, in most cases), and because machine learning and artificial intelligence are becoming affordable and accessible talent acquisition solutions, none of these things are are actually a sustainable, scaleable strategy for recruiting success.
It’s probably a bit premature to start penning that obituary for recruiters just yet. Because as cool as using a Chrome extension to find out how a candidate stacks up against the competition or what their contact information might be, this whole concept of “algorithmic sourcing” – which is, programmatically capturing and converting leads online – may already be here, but it misses the point.
While digital marketers have relied on these sorts of tools and technologies for years, now, most digital marketing – algorithmic or otherwise – has been complete and total crap, somewhere just south of awful (at least to date).
Sure, it’s a longstanding trend that recruiting and digital marketing are essentially analogue functions, and that we should be approaching our candidates the same way our marketing colleagues have long approached consumers, but based on what I see in my inbox every week, this conventional wisdom is, in fact, total bullshit.
We haven’t actually advanced through the adoption of algorithms – if anything, we’ve actually devolved into some sort of digital dystopia where marketing emails now look like this.
I haven’t heard back from you and that tells me one of three things:
1) You’ve already chosen a different company for this, and if that’s the case please let me know so I can stop bothering you.
2) You’re still interested but haven’t had the time to get back to me yet.
3) You’ve fallen and can’t get up – in that case let me know and I’ll call 911
Please let me know which one it is because I’m starting to worry…Thanks in advance and looking forward to hearing from you.
Some Douchebag in Software Sales.
If this sounds familiar to you, then maybe it might be time to stop listening to all this talk about AI and machine learning, to cease ceding that algorithms can best predict when someone will move AND when they’ll accept an offer and realize that this is more product marketing than possibility. I mean, seriously.
We can’t even seem to develop an ATS that allows for searching a database more accurately than simply stack ranking results based on keyword density, or return more relevant results than your average Boolean string, and yet, all of the sudden, we’re ready to hand the keys over to the same legacy vendors who can’t even develop mobile optimized career sites or accurately measure source of hire?
If you can’t even get the basics right, I’d probably say we’ve got some time before we should really worry too much about being replaced by robots.
Some Rambling Notes on An Idle Excursion.
Let me be very clear: I think most recruiters actually like technology, and are NOT adverse when it comes to adoption. I, for one, see a marked value when it comes to the application of artificial intelligence and machine learning to talent acquisition. But I think that we live in an age that’s about equivalent to having to crank your own car, as far as recruiting technology is concerned.
We’re not all that far removed from the days of faxing resumes to our clients and having to print out offer letters and mail out hard copies to our candidates. We have a hell of a long way to go before we have anything even resembling a recruiting system that completely replaces people, if we ever hit that point. And I’m dubious that we ever will, to be honest with you.
What’s really maddening to me is the fact that many of the same pundits and product marketers so actively evangelizing artificial intelligence and machine learning (presumably, mostly for great click through rates, social sharing and “influence” – gag) are the same ones so willing to scorch the earth with the type of canned spam that’s made an outhouse of our inboxes these past few years.
Let’s go back to 2010 – I know it seems like forever ago, but even back then, after six years on the market, recruiters are still “InMauling” candidates like they were swapping partners at a Roman orgy. These generic messages are pervasive, intrusive and 100% impersonal – and most are terribly written, just to add insult to injury. But man, it’s suddenly easier to spray and pray than, well, at a Roman orgy.
And over the intervening few years, we shoot so many empty loads that eventually, our message not only becomes impotent, but off-putting. The roar of pissed off candidates becomes deafening, and suddenly, our industry dialogue shifts from high tech to high touch, and we begin adopting a much more personalized, much more targeted method of messaging and engaging our candidates.
What Is Man (And Other Observations).
This reversion to personalization and relationship building means that suddenly, it’s not about who you can source with those Boolean strings, and competitive advantage is no longer being able to source lists for email blasts and automated outreach.
Instead, researching candidates has gone from demand gen to lead nurturing, and recruiters are finally realizing success requires taking the time to find out the kinds of personalized information and relevant messaging necessary to presumably (and ironically) prove that we’re real people and not automatons and that our professional services aren’t just another beat up piece of SaaS.
Of course, after that initial outreach, the algorithms that identify candidates and maybe even predict future behaviors as individuals by analyzing historical data in aggregate will never supplant what has become the most important function of our function: engagement.
At the end of the day, engagement is everything, and interpersonal relationships are king – content be damned. If you can identify candidates but don’t know what they want and how to get them to trust you, then all of the artificial intelligence and machine learning solutions in the world can’t help you with hiring.
Think about it: as a species, we’re just getting used to the fact that self-driving cars or using our fingerprints to unlock our phones are more fact than science fiction, and yet all of the sudden (or within the next two years, as many posts would have you believe) we’re going to take a process that’s not only one of the most important parts of our professional existence but is one of the top three most stress inducing activities in our personal lives, too, and suddenly hand it over to a machine?
I mean, come on, the same people who predict artificial intelligence in recruiting is going to bring down the entire industry are the same exact people who bitch and moan when they can’t speak to a real human when they call a customer service number, the same people who still write checks because they don’t trust PayPal and the same ones who still go to networking events and trade shows because “there’s nothing like being face to face.”
They’re not prepared to let machines do the work anymore than they are to toss that whole “candidate experience” thing to the side, either. Technology can only take you so far, but until it can take you through an application process, then machines have their intrinsic limitations.
Sketches New and Old: History Repeats Itself.
For a decade, we’ve spent an inordinate amount of time and effort clamoring about how important the “Candidate Experience” is, so much so that we’ve even built award programs and “best practices,” not to mention a codified discipline complete with dedicated “candidate experience specialists” around something that’s pretty much just common sense and UI/UX.
And even with our collective embrace of “personalized outreach” and “job seeker engagement” or whatever you want to call what’s more or less just the Golden Rule, emails like this still go to the CTO of Amazon, the guy who basically invented cloud computing.
If you cast a wide enough net, you’re going to trap yourself sooner or later, right?
In recruiting, we’ve spent a ton of time learning not only how we can leverage technology to create efficiencies and increase effectiveness, but we need to shift that same focus towards ensuring that we’re getting the fundamentals down, first, and that any talent technology augments and enhances interpersonal communication and engagement rather than purportedly replaces it.
Recruiting comes down to making a connection with another person and building enough trust to convince them to consider your company or opportunity – machines can tell you which candidates are the likeliest to consider making a move, but having an offer accepted (and a requisition closed) is really the end goal of all of this.
And no machine on earth can credibly allay a candidate’s concerns, be savvy enough to negotiate an offer with both sides feeling like they walked away winning or provide the hand holding and coaching required to get a candidate to actually go through with exiting their existing employers and coming onboard at your organization. Machines can influence outcomes, but ultimately, they don’t make hiring decisions, and they don’t accept offers – humans do.
Which is why recruiters, contrary to popular belief, can never truly be replaced. We’re the ones converting candidates into hires – often in spite of our tech and tools, not because of it.
On The Decay of the Art of Lying.
One of the biggest complaints most recruiters make when actually using recruiting technology is that while these suites and systems are built for recruiters, they’re rarely, if ever, built by actual practitioners who have any use for the end users who have to live with their products on a daily basis. Think about it.
If some great artificial intelligence system is going to predict which of your employees is the likeliest to make a move, when they’re probably the most poachable and about how much they’re likely to get offered for their next opportunity.
Do you think recruiters are the ones responsible for hard coding these algorithms and engineering the real variables involved in real recruiting to ensure that these systems can somehow replicate the lessons learned over a lifetime in the talent trenches?
If your answer was a resounding “yes,” then congratulations, what you just said is one of the most insanely idiotic things I have heard.
At no point in your rambling, incoherent response were you even close to anything that could be considered a rationale thought. Everyone in the room is now dumber for having listened to it. I award you no points, and may God have mercy on your soul.
The truth of the matter is that it’s only been in the past few years that many emerging and established HR Technology providers have come to the realization that market research (or “voice of the customer,” if you want to sound cool) really matters, and if you’re positioning your technology as a “solution,” you’d better damn well know what the problem if, first.
Many of these vendors have begun assembling seasoned practitioners on their product development teams and advisory boards of actual talent leaders to help build a product that meets their needs instead of one that just checks all the right boxes on an RFP. The ones with this approach to R&D almost inevitably have a distinct advantage over those built by coders, engineers and techies with limited exposure to recruiting and no experience in talent attraction and selection to speak of.
Those are the products whose roadmaps seem to lead inevitably into oblivion.
The Prince and the Pauper.
So the next time you wonder when the machines are going to be taking over and rendering your recruiting department obsolete, take a step back and ask yourself a question: can artificial intelligence know, or machines learn, not only who’s likely to make a move, but get at the heart of why they’re looking and what they really want out of their next opportunity?
Can they determine that a marriage is crumbling because of work related stress, having to work so many hours that they barely see their family or that their only non-negotiable has nothing to do with title or money, but instead the chance to spend some more time with their spouse and their kids ?
Will they know exactly what that person’s manager is doing to make them so miserable, or how under-appreciated and overwhelmed they feel in their current roles? I seriously doubt it, but a great recruiter who takes the time to get to know and understand a candidate can figure out fit far better than any algorithm ever could.
With the automation revolution in recruiting already firmly underway, we should be seeing an impact on results and improved efficiencies and efficacies when it comes to talent acquisition and hiring, right?
I mean, since the seismic shift towards AI and ML seems to be more of a foregone conclusion than a speculative possibility – which is what the “influencers” would have you believe – then you’d think we’d already be writing an obituary for the recruiting profession instead of hiring them at historically high levels.
But employers seem unable to meet the insatiable demand most businesses seem to have to help solve the supply and demand issues endemic in such a cutthroat and competitive market for top talent, because like any other highly skilled role, good recruiters are hard to find.
Reports of our death have been greatly exaggerated, but if you’re a recruiter who actually recruits, you already knew that.
This article first appeared at RecruitingDaily on Dec 7, 2016.
The day a machine can recruit humans is the day we don’t need humans to do the job it’s recruiting for.
Ivan, I see your point, but while w may be able to teach the machine to source and find profiles, we cannot teach it engagement. There’s a human component that will always be a part of the process. Machines have no empathy.