I try to not to get all worked up over recruiters being replaced by robots, or recruiting-is-dead click-bait articles, but some of them are just annoying as heck. I want to try and bring some additional facts and sanity to this discussion, as most of the discussions I have seen are based on emotional and anecdotal responses.
I also wanted to bring other industry voices into this discussion like China Gorman, Gerry Crispin, Kevin Wheeler, and Glen Cathey as we believe this particular topic has an enormous impact on the future of our profession.
“This article is fascinating! You build a great case for recruiters to take seriously the advancements that AI, algorithms, bots, and VR are making in the talent acquisition space and tie it up in a way that makes one look away at their own peril! — China Gorman
I am not a futurist. I don’t have a crystal ball, but what I do have is 20 years of global recruiting experience leading functions all around the world, and I have been an agency and corporate recruiter. I have also spoken to thousands of recruiters, HR, business, and talent acquisition leaders in my career. Most recently in my current role, I have looked at thousands of survey results on a broad array of recruiting topics over many years. Did I speak with or survey every recruiter on the face of the earth? Of course not. That should not lessen what I have to say next; I hope not.
The following is my opinion from all these conversations and all the balanced research I could get my hands on related to the subject. Flame me, love me, hate me, I just want to share some interesting observations on the recruiter vs. robot debate, and no recruiting is not dead, but it’s sure not well either.
Based on all of these observations, people in recruiting break out into my graphic above when it comes to change. Seventy percent don’t care to change and/or are just downright clueless.
So what I say here next and following might brandish me a heretic in HR circles.
The majority of you in recruiting are going to be redundant and you don’t even realize it. (click to tweet: The majority in recruiting are going to be redundant and don’t realize it. )
There are two things I have noticed accelerating this point of view in recent years that are starting to converge:
- High levels of dissatisfaction by candidates and the business
- Rapid increases in technology innovation and convergence of technologies
Recruiting is meant to be a people business. When the people in recruiting don’t add value, don’t treat another human being like human beings, when such a large majority of recruiters don’t care or are just clueless, then you’re opening the door to change and disruption.
Try this search in Google….
To be frank, clearly I think that we all know that the majority of our so-called recruiting brethren are not liked, trusted, or just downright despised. While I am sure many of you don’t fall into this 70 percent bucket and do add value, to tell you the truth, candidates, and hiring managers, on the whole, are begging for something different.
“Change happens when the pain of staying the same is greater than the pain of change” —Tony Robbins
So here we are today.
Now let me address my second point.
Very few of us can sit here with confidence today and say that the assessment of candidates against job requirements has become an exact science with high levels of predictive accuracy. Is this 100 percent on the recruiter’s shoulders? No. We could all run through a litany of reasons why from poor job definitions of what defines key success criteria on the job through to hiring managers asking asinine questions like “how many ping-pong balls can you fit in a 747?”
Treating candidates like human beings during the interview process is just as much, if not more so, on the business’s shoulders, as you can see confirmed from my Google search above.
The current ménage à trois between hiring managers, recruiters, and candidates is incredibly dysfunctional, with each party constantly pointing the fingers at the others for their levels of dissatisfaction. But unfortunately for the recruiters in this three-way relationship, they are the ones with the least amount of power. The don’t make any hiring or buying decisions. They influence at best.
So right now one of the core things that recruiting has going for it is no one has cracked the code on technology doing a better job of assessing performance than a human being, regardless if the majority of us humans still do a crappy job of assessing talent today anyway. Also, most hiring managers don’t treat the identification and assessment of talent a priority in their job responsibilities, as it takes up too much time and/or they don’t get why it’s important. That means, while it’s too much of a pain in the ass for a hiring manager, they prefer that someone else do it. Remember, this is not all hiring managers, but I think if you bear some recruiting battle scars, and you’re honest here, you will probably err on the side of the majority.
Can a robot (algorithm) do a better job of finding and assessing talent? That’s the real elephant in the room. (click to tweet: Can a robot (algorithm) do a better job of finding and assessing talent?)
In the future, yep.
Who knows, but to say it’s not going to happen is akin to people saying years ago that self-driving cars are a fantasy. Read on and I will explain why I make that statement.
As you continue to read don’t think in terms of physical robots like Robby (no pun intended), C-3PO, The Terminator or automatons, but rather, think in terms of software that performs a task.
Let me ground you in something that I think is very important first. Regardless if you are great at your job or suck at it as a recruiter, face the facts …
Recruiters Are in a Middle-person Business
Yes, I know that a lot of you on the corporate side do other things like employment branding, optimizing the candidate’s experience, etc., but if we really boil it all down, the role of the recruiter is to identify, attract, and assess talent.
If history had taught us anything, companies that are in a middle person business that doesn’t add value or disruptive technologies come along, then they are banished to the history books.
Remember Blockbuster brick-and-mortar stores? I can go straight to my TV now and cut out the middle person. Heard about Uber and its interest in self-driving cars — no impact on taxi drivers, right?
What I am not talking about is the total inhalation of the recruiting species by Skynet. I am in the camp that technology might destroy some jobs while creating new ones. What I am saying, though, is a lot of the job description and tasks of the everyday recruiter does today will be gone. I think it’s a not if, but a when question.
“Machines and computers don’t perform jobs, they automate tasks.” — Jerry Kaplan, Fellow CodeX – The Center for Legal Informatics, Stanford University, from his talk “Humans need not apply” at the Google campus in 2015
Sorry folks, it’s not about purple squirrels and unicorns.
If you look at source of hire surveys (which we know are dubious anyway), and even if you doubt those surveys but have been in recruiting for all of five minutes, what you will notice is that still today 80 percent+ of the hires made by the majority of companies are from candidates coming to you: existing employees (internals), employee referrals, your own ATS, contractor conversions, job postings, job aggregators, your own careers page, social media pages, organic search, and paid for search (PPC).
Even if surveys are dubious in their accuracy, and we just went off directional correctness here, you will be hard pressed to find people disputing that the majority of companies don’t get the majority of the hires from these channels.
The point I am making and before you start banging a curt response on your keyboard, Ok, I know what you are thinking. Yes, there are still plenty of positions where we have to go find the candidates vs expecting that you can post the job and sit back and wait. Based on my own experience, surveys (directional remember) and conversations with many others, on average about 10-20 percent of a company’s hires require a proactive, you-go-and-find-and-attract-them strategy still. Once again, exceptions yes, but the majority rule is most companies hires still come from the 80 percent+ who come to them.
Think about that for a moment and let it sink in. That means when technology reaches a point to do a better job of the assessment of a candidate, a large number of recruiters are gone given the reality of the 80 percent+.
Hey, where did Waldo the Recruiter go?
Let’s go back to the Tony Robbins Quote. Change happens when the pain of staying the same is greater than the pain of change. Right now today as I type I know of many HR tech companies working on trying to come up with a smarter solution to help both candidates and hiring managers determine a better fit. Yes, I have seen glimpses of this already (see below) but we are still not quite there yet.
In short, the pain to change I don’t think is the issue, but rather there is no real scalable or broadly universal solution that gets to a better outcome today while at the same time providing an optimal user experience. People don’t want to answer 100+ questions on a matching solution like they are on some online dating site. They want a quick drive through, “can I have a side of fries with my job application please?”
Of course in the future, at least in the short term we all have nothing to worry about. Even in the medium to long term, there will always be a home for great recruiters as I can’t imagine any technology at this stage that can proactively find and attract all the talent who are not looking. Never say never, but I can’t grasp it right now. But when it comes to the assessment side of the equation, then this one is like a light at the end of the tunnel, and fellow recruiters that light that keeps getting closer is a disruptive train on a track ready to run you over.
You the reader fall into two camps on this at this point. No f’ing way, as what recruiters do require skills that no computer can never replace. To be fair, most of my career I fell into the no f’ing way camp too. As I have gotten older I have seen steady advancements in technology while continuing to see too many recruiters not really change or just continue to provide such a shitty service to candidates and hiring managers.
Back to my original pain point about change. You will have very little resistance from candidates and hiring managers given the current experience (or lack thereof) once someone cracks the code on technology effectively assessing the candidate/job/company fit.
At this point if you are one of those people who thinks everything I have said so far is crap, then you can hit the back button now. Don’t worry, I will not be offended. If you fall into the “well, maybe” camp, then let me give you some interesting insights if you are not aware some of the technology that exists today. While some of these technologies are not mainstream yet, or within the financial reach of mainstream recruiting, they are going to be sooner or later. If you believe (or don’t) in Moore’s Law or Rock’s Law, then the speed of this change is surprising some experts (aka AlphaGo, see below)
What we are seeing in the advancement of computer AI, algorithms, and deep machine learning, makes you not only think what is possible and what is next but shows you what is achievable today.
What about a robot doing a better job at assessing candidates than a human?
Meet Sophie the Robot back in 2013. Asking 76 questions about selling, recording emotional responses and facial expressions by candidates.
“She captures their [candidates] cognitive verbal responses and captures their emotional responses by monitoring changes in their facial expression, Khosla says. The whole idea is to develop the emotional profile of candidates including their passion for the job and a behavior profile and benchmark this against an organization’s best employees. Khosla insists robots will not replace humans conducting later stage interviews or employers making final hiring decisions.”
Makes you think about Khosla’s last statement. “Robots will not replace humans conducting later stage interviews or final hiring decisions.” Last time I checked the majority of recruiters do top of the funnel pre-screen interviews and don’t make hiring decisions.
In 2016, AlphaGo beat the world’s best player in the game of Go.
“So what,” you say?
This is a very big deal in the advancement of technology paralleling the complexities of the human brain. If you are not familiar with the game of Go (I was not), it is an ancient Chinese game that has more moves possible than atoms in the universe (Google Quote). It requires a totally different type of thinking that people did not think was going to be achieved by a computer for many, many years. In context, IBM’s Watson computer beating Gary Kasparov in chess or even beating the world champ of Jeopardy dwarfs in comparison of what AlphaGo achieved.
Google acquired the UK company DeepMind in 2014 for 400 million pounds. (DeepMind is the creator of the computer AlphaGo). Here is the Wired Article if you are interested in what impact this has in the advancement of AI, but in short if you could not be bothered to read the full article, this statement gives you some insight:
“It marked the first time a machine had beaten the very best at this ancient and enormously complex game — a feat that, until recently, experts didn’t expect would happen for another 10 years.”
AlphaGo algorithm is structured differently to traditional computers. It uses a neural network of large amounts of data (experience) in a combination of deep reinforcement self-learning algorithms. It short, it uses intuition in how it leans and works which historically has just been the domain of us, humans.
Think about Sophie the robot in 2013 but now with the possibilities of thousands of interview questions (competency, behavioral, situational, functional, technical, etc.) as the neural network with reinforcement self-learning algorithms (employee performance, engagement, and retention data) to cross reference better outcomes. This does not exist today of course, but who knows, maybe Google is already working on this in-house itself.
I can’t give an explanation of the algorithm the justice it deserves, so if you want to geek out and understand how it works, you can watch this video from the CEO of DeepMind.
The Future of Artificial Intelligence
By the way, the algorithm is also open source code now for others to build upon, which is only going to accelerate innovation.
Companies Are Using Algorithms to Assess and Retain
Xerox Article — When looking for workers to staff its call centers, Xerox Corp. used to pay lots of attention to applicants who had done the job before. Then, a computer program told the printer and outsourcing company that experience doesn’t matter.
“Laszlo Bock, a senior vice president at Google Inc. and an Evolv director, said the software will supplement, if not supplant, many of the personnel decisions long made by instinct and intuition.”
Quartz Article: Algorithms make better hiring decisions than humans.
Managers might like to believe that they have better-hiring judgment than a computer, but a recent working paper (paywall) from the National Bureau of Economic Research suggests otherwise.
The researchers looked at the employment record of 300,000 low-skill service sector workers across 15 companies. The jobs had low retention rates, with the average worker lasting just 99 days, but researchers found that employees stayed in the job 15 percent longer when an algorithm was used to judge their employability.
Can the Marriage of Watson and Amelia Produce the Perfect Robot?
Today IPsoft Amelia and IBM Watson’s cognitive technology come close to passing the Turing Test. It’s virtually impossible to tell the difference between Amelia, IPsoft’s virtual assistant, and a real live person on the other end of the phone. Not only can she understand what people ask in 20 languages, she also responds appropriately to the caller’s emotions both visually and verbally. All that aside, here’s where Amelia and Watson fail: they struggle with simulating our less intelligent behaviors. They don’t make enough mistakes and they don’t lie well. Not to worry, though, they’re working on that.
A seeing, reading, writing, and hearing Algorithm. Hey, don’t humans do that?
“80 percent of the world’s employment in the developed world is the stuff computers just learned how to do.”
Jeremy Howard TED Talk, the wonderful and terrifying implications of computers that can learn:
This is now getting serious …
One of the largest law firms in the world, Littler Mendelson, who also happen to be the first law firm to create a practice area of robots, AI, and automation in the workplace. Here are a few excerpts from a Littler Mandelson’s PDF report on Robotics in the workplace, plus the Pew Report on AI, Robotics, and the Future of Jobs
“One of the most advanced systems features Arya, identified as the first recruiting robot. Arya is virtual and learns the search patterns used by the recruiter it is assisting. Arya then scans the web selecting candidates and independently arranging interviews. All the workplace laws that govern the way employers consider applicants apply equally to this existing technology that makes candidate selection decisions.
“In sum, robotics is one of the fastest-growing industries in the world. It has been estimated that by 2025, half of the jobs in the United States will be performed by brilliant machines and intelligent systems. Robotics is the next major innovation to transform the workplace, and will have as great — if not greater — impact on how employers operate than the Internet.”
Luddite, Transhumanist, or Pollyanna … pick any two.
I see many article and blog comments about what the future recruiter role looks like. Heck, I don’t know, and for people to claim unequivocally the do, is BS. I also don’t like cliché statements like “it will be different than it is today” is a cop out. I can guess, but it will be just that. If you put a gun to my head and say “guess Rob,” then here’s my best:
- There will be significantly less of you/us. What is significant? I’m guessing up to 90 percent. There will always be a role for the good recruiters proactively finding and attracting talent, but that number is going to get smaller and smaller in the future.
- Some smart recruiters who have both analytical and operational competencies will take up program management roles of optimizing and managing the overall talent attraction and assessment strategy for a company. Looking for data insights and knowing how to adjust the company’s strategy to gain a competitive advantage.
Beyond those two, I would have to leave that up to people way smarter than me. If you’re interested in the AI topic in general and the impact it potentially has on all our futures (beyond just recruiting), then here are some interesting thoughts on the topic.
Are Droids Taking our Jobs?
- Rise of the Robots: Technology and the Threat of a Jobless Future
- Our Final Invention: Artificial Intelligence and the End of the Human Era
- The Second Machine Age: Work, Progress, and Prosperity in a Time of Brilliant Technologies
Am I rowing in this boat alone?
As I stated above, there are a lot smarter people than me in our profession, who have seen more than I have seen so I shared my article with them for their comments and observations as well.
Note: I did not ask them to support my observations, but rather give their own opinions based on their own perspectives.
Gerry Crispin, Principal & Co-Founder, CareerXroads comments …
“I have the advantage of being able to ignore 90+ percent of the recruiters who view recruiting as a job and no more — a job to be done with as little effort as can be viewed as acceptable and with as little personal investment in or accountability for a quality outcome. I may ignore this ‘majority’ but I’m still aware of them.
What keeps me in the game and optimistic is the core community (10 percent) of recruiting ‘professionals’ who actually see the work of recruiting as a profession or trade to be proud of and, where their learning and growth are ongoing — by following the job seekers they’ve helped hire as they morph into competent and contributing employees. They strive to see the entire supply chain in its various modes and are daily compelled to improve their skill and gain experience as knowledge is acquired from data connected & analyzed, decisions made, and results shared. They embrace change, learn continuously, and step up as an equal when working with their ‘clients’ by bringing market data, buy or build challenges, and much more to the conversation.
Assessment/selection is but one of the broken elements of current recruiting models. I would argue however as others do that the future is here — just not well distributed important and even critical as selection is, the science [approach] of every facet of recruiting, selection included, has already been cracked. The gap between the present and the future isn’t technology but merely the improvement in the collection, cleaning, and application of the data that now abounds. As it gets more practical and less expensive to just do it … we will.
We may differ on nuance but, I absolutely agree with the premise that, at some point in the not-so-distant future, an intelligent automation that mirrors nearly every aspect of today’s recruiter actions … and then some — from the initial point of a candidate’s awareness of the firm as a potential employer to the last vestige of onboarding, can all be done significantly better with technology alone- no human.
If I had, as an example, the choice of maintaining 40 recruiters to help hire 5,000 hourly employees in a given job family over the course of one year versus tweaking ONE artificial intelligence system that senses need, deploys messages targeted to the appropriate audience, responds as if it were a caring human, screens, schedules, checks for problems, offers appropriate rewards, and takes care of pre-boarding — well, it’s just inevitable which decision would be right for the business’s time, cost, and quality. Only hourly you say? No, not only. Just first.
Ninety percent of the people responsible for recruiting will disappear. Ten percent will remain to design, manage, and build strategies around the AI … because they understand the pieces of the puzzle and that what is under the hood is still an ever-changing set of ideas to improve, compete & innovate — then execute at a world-class level to acquire talent.
I’ve no doubt which few recruiters will remain.
Glen Cathey Comments …
“Can a robot (algorithm) do a better job of finding and assessing talent than a human? Perhaps — but it is critical to ask what constitutes “better.” Slightly improved retention? I think that’s a poor indication of “better” — should we be impressed because people stay longer? Without fully opening that can of worms, who’s to say that the people who leave aren’t “better,” and that they leave because they have more potential than the employer recognizes? I’d be more impressed when companies can report that automation in the hiring process has improved revenue per employee.
The “future” is already here — assessing, selecting, and hiring can be and already has been largely automated today. Just take a look at what SAP is doing in this space. Granted, most of the examples of successfully leveraging automation in the hiring process focus on low-skill or entry level jobs (which may be automated themselves in the very near future!), but that’s not to say that the recruitment for higher-end jobs can’t be similarly automated at some point. Given enough time, investment, and ingenuity, very few things are impossible.
Automation of the sourcing process comes with its own set of challenges. Using job descriptions and previous successful hires to automatically create searches to find potentially qualified candidates in ATS’s and online job boards might sound like a dream come true for some companies, but job descriptions and resumes are inherently flawed.
With tight controls on title and skill matching, this kind of automated approach can do as good of a job as an average human at finding potential matches. However, this level of searching produces false positives and excludes a great many potentially viable candidates.
Using automation to reach out to the sourced candidates to gauge their interest in the position is essentially email blasting. Even with tight filtering, there’s going to be a good portion of people who will get irrelevant messaging — it will essentially automate what some recruiters do today (run title and keyword searches and send emails to everyone in the search results). With effective messaging/candidate engagement and #fightspam top of mind in the talent acquisition community, automation of what isn’t already working particularly well isn’t a step in the right direction.
Furthermore, if the majority of “recruiting” as we know it today were to become largely automated, companies would pretty much all be competing in the same shallow end of the talent pool — the ~20 percent of people who will take action to apply to jobs and respond to digital marketing efforts. Wait a minute … is that really any different than today? 🙂
Now, unless you’ve been living under a rock, candidate experience is all the rage these days. One does have to wonder what the use of algorithms and automated assessments might do to the candidate experience. I’ve wondered this myself, and I’d like to ask you if you think automation has improved something as simple as customer experience when it comes to calling banks, credit card companies, phone companies, customer service numbers, etc.?
Regardless, the ultimate question is whether or not companies will care or be materially impacted if their use of algorithms to find candidates only skims the surface of the talent pool and spams volumes of people, or if their use of automated assessments delivers a less-than-optimal candidate experience as long as it “works” (results in filling roles with qualified candidates).
At some point, however, with increases in the adoption of automation in the hiring process, I can see some companies decidedly NOT using automation as a competitive differentiator (“speak with a REAL person!”). Imagine that.
While automation strikes fear into the heart of talent acquisition professionals, let’s not miss the fact that AI is already impacting the legal and medical fields, populated with highly compensated professionals presumably performing highly complex and knowledge/judgment-critical jobs. Algorithms have already proven they can do a better job at legal research and medical diagnosis and appropriate recommended treatment. Why would we expect sourcing, recruiting, screening, and selection be immune to automation not only replacing but improving upon humans?
I’ve already argued (unpopularly) that some folks today who have the title of “recruiter” aren’t actually recruiting people — they are processing applicants. When you put it that way, is it really that difficult to imagine automation displacing the average recruiter at some point? The role of the loan processor has already been reduced through the use of algorithms — humans exist for exception handling. If a human isn’t adding value to the process, why not automate it if the end result is essentially the same, if not better. (Again, “better” is a slippery slope …)
Granted, there are recruiters who aren’t just processing inbound applicants — those who perform outbound recruiting, proactively sourcing the deeper end of the talent pool (~80 percent of the total potential candidate population), converting people into viable candidates who would otherwise never become applicants, regardless of how fancy and impressive a company’s employer branding and digital marketing efforts may be. This kind of outbound recruiting activity is remarkably similar to the traditional sales process, and I don’t think I’ve seen much in the way of algorithms replacing salespeople. In fact, I always like to point out that over the past two years, Facebook has about the same or more openings posted online for salespeople as they do software engineers. As George Anders points out, “Even in the highly automated world of online marketing, it turns out that making a deal come together still requires a human touch.” Chew on that for a bit.
I’d like to leave readers with these (hopefully thought-provoking) questions:
- Are we looking to technology to solve problems in the talent acquisition process, or are we looking to identify effective solutions to the existing problems in the hiring process and then apply technology to support people in the process?
- Is our drive for automation in the talent acquisition process a crutch to avoid having to think deeply about improving the existing process?
- Will automation technologies replace people or instead support people doing the work in the continuous improvement of the talent acquisition process?
- Do we believe that having the latest, fastest, most sophisticated technology is necessary to maintain a competitive advantage when it comes to attracting and acquiring talent?
- Is there a danger in not paying specific attention to the fact that people aren’t creating talent acquisition automation solutions for altruistic reasons? (hint – they see $$$, and just because they sell it, it doesn’t mean you need it)
- Will the introduction of automation into the hiring process help eliminate waste, as per Lean? (anything that does not add value is waste – defects, inventory, over-processing, over-production, waiting, motion, transportation).
- Just because we can automate some or all parts of the talent acquisition process, should we?
Kevin Wheeler, president, Future of Talent Institute comments …
“The location and assessment of candidates, as well as much of the engagement process, can and will be taken over or significantly augmented by automation and robotics. The software that will make this possible includes all sorts of passive assessment tools, as well as chatbots and other engagement tools. Most of them exist today. New ones are emerging as we read this. It is a growing market that is getting more sophisticated every month.
The questions that remain is whether or not there will be a need for recruiters. And I think the answer is yes. The world tried to remove the travel agent through automation and did, indeed, reduce the number. On the other hand, it gave new life to the agents capable of putting together complex trips and those who could maintain a relationship with corporate clients. Today these agents are thriving and use the automated software to augment their skills. I imagine the same will happen to recruiting. As automation has taken over the booking of simple trips and makes it easy to compare prices, the same will happen with recruiting. Candidates will have information on salaries and other issues that are not usually available today. Anyone looking for or recruiting for simple jobs, routine jobs, or hourly jobs will find the experience totally automated. More complex jobs and highly skilled jobs will still be fulfilled by a recruiter, but a much different one than we have today,
In a few years, the run-of-the-mill sourcer, transactional recruiter, etc. will be largely history. They will be replaced with recruiters with superb relationship and influencing skills and who are comfortable leveraging the tools that automate the rest of the process. These people will be the interface between hiring managers and candidates with the interests of both in mind. Their job will be to mediate the best possible match. They will not work inside corporations. Indeed, I see the growth of RPO and a significant decline in the number of in-house recruiters. The data seems to support this as RPOs have been growing at major rates and the ones who provide a quality experience are growing even faster, there is less and less need for corporate recruiters.
And the train went Choo! Choo!
For those of you that stayed with me during this article, thank you. I hope it encourages discussion and debate. The reality is that if you are reading this then by default you are probably in the 30 percent who understand how to add value or wants to learn how and regardless of what the future holds, are willing to change, adapt and get better.
What I hope for the most, though, more than anything else, is maybe articles like this becomes a wake-up call for the 70 percent. Either start adding value now or get run over by the ensuing train.