Why Recruiters Resist Automation
And Why They Shouldn't
 |
Is automation the inevitable future of recruitment? Will artificial intelligence replace the recruiter?
These questions are asked frequently, and the opinions on both sides bounce around the Internet almost daily without resolution. Recruiters are more resistant to automation than most other workers, perhaps because they work primarily with people and not with processes and machines. They are too close to the problem and have too much invested in maintaining their current state to be useful critics or innovators. And, like everyone else, they are fearful of losing their jobs.
Their arguments are basically of two types. They argue that it is impossible to automate many recruiting functions such as sourcing, assessment, and engagement. They are too complex and involve understanding the nuances of a hiring manager and the corporate culture. Automated tools cannot do these with quality. They also argue that recruiting is a personal affair, and candidates expect a face-to-face encounter. On top of that, automated tools are biased and cannot be trusted to make objective decisions.
There is truth in these arguments, but they do not fully account for the benefits these tools can provide them or the continuous progress developers make in improving these tools.
Users of a product or those who provide a service often can describe a problem, but rarely do they come up with the best solution or the solution that would benefit them the most. As Henry Ford once said, “If I had asked people what they wanted, they would have said faster horses.”
Even though recruiters complain about having too many requisitions to fill, the talent shortage, fickle hiring managers, and legal hurdles, they still do not appreciate that automation and A.I. could free them to do more useful work.
This is not unusual. For example, tellers and many bankers thought that ATMs were just a nice addition to their services and never considered that they would replace bank tellers and automate most banking freeing up people to become loan officers or financial consultants.
So it is natural that recruiters have a hard time accepting that recruiting may be largely automated and that candidates may actually like an automated process and find it better than working with a person.
History is also not on their side. History tells us that in the end, there will most likely be a partnership between people and A.I., with A.I. doing most of the tedious and routine activities involved in hiring and recruiters taking on roles that involve personal relationships.
In failing to embrace automation, recruiters make many assumptions without the data or experimental evidence to prove or disprove them.
#1. Automation is impersonal and bad for the candidate.
What does the evidence tell us?
Automation is not bad in itself. It can improve the recruiting experience by making it less bureaucratic, faster, and more objective. But it does require recruiters to adapt and figure out how they can add personalization and communication with candidates and hiring managers. Rather than being stressed with administrivia, they can devote more time to developing and nurturing relationships. And, after all, this is the essence of good recruiting. Almost everything else recruiters do is wasteful.
#2. They assume that people prefer to communicate and get information directly from other people.
What does the evidence tell us?
Chatbots are popular, and data shows that candidates like them for their speed and ability to provide information and answer questions without the hassle of trying to connect with an elusive recruiter. Likewise, FAQs and other data sources are heavily used.
#3. They believe their own intuition and experience are better predictors of success than tests or other assessment tools.
What does the evidence tell us?
We have tons of academic research that show how inaccurate and biased people are in deciding who to hire. Charles Handler, a leading I/O psychologist, recently wrote an article in ERE.net that discusses this in detail. One of my recent newsletters also talks about how invalid our interviewing process is.
#4. They believe they are needed to influence candidates, coach them, and also influence hiring managers.
What does the evidence tell us?
There may be some validity to this assumption, but automation does not exclude them from coaching candidates or hiring managers. Overall, candidates are most heavily influenced by the organization’s reputation, products, services, level of innovation, benefits, salary, and location.
#5. They believe they have insights into the hiring manager's needs and personality and choose candidates that the hiring manager will want to hire.
What does the evidence tell us?
Artificial intelligence, combined with analytics, can give recruiters tons of information about who a hiring manager has hired, what the consistent traits are that appeal to them, and much more. A smart recruiter would embrace A.I. and analytics to be even more effective in influencing managers and in choosing the right candidates to present.
Automation offers to make recruiting a truly candidate-centric and personalized experience and frees a recruiter to spend more time adding real value by coaching, using data to help select better candidates, and simplifying the entire process.
Subscribe
Thanks to all that have recently subscribed. It is greatly appreciated and helps us continue to research and write these unsponsored articles. We try very hard to be neutral and objective. A subscription is only $30 for an entire year. As our paid subscriber base grows, so will the content.
Your upgrade to a paid subscription will give you access to various special reports, white papers, videos, and more. We have recently published a white paper on the future of RPO, a series of links to key readings on A.I., and an exclusive interview with Felix Wetzel, Director of Career Acquisition at Cera. Much more coming soon.
Related Links
| 2 | 4 |
Always delightful to read your pragmatic views. Like you, I believe that AI as a substrate opens up new portals and insights related to things such as labor and the talent economy - but here's the twist, what if there isn't an actual "talent shortage" as you pointed out? What evidence do we really have for that? Furthermore, what will AI tell us in the 2020s about our current talent landscape? My guess is that automation will reveal talent surplus, not scarcity. I am open to discussing this further and look forward to your thoughts. Thank you, Brad.
In 1966, in the Our Man Flint spy-fi, comedy/farce/parody trilogy of the James Bond franchise, the US govt agency Z.O.W.I.E. (Zonal Organization World Intelligence Espionage) was faced with dire circumstances. Leadership was at a loss on how to respond. They needed to know what agent talent could save the world.
Fortunately, at Z.O.W.I.E. HQ, they had an IBM Type 75, Model 2, Card Sorter. The director, surrounded by top military brass, walked over to the Type 75. They waited patiently. Then, the Model 2, Card Sorter dropped a Hollerith punched card into a slot. The Z.O.W.I.E. director looked at the card. Although somewhat dismayed, they had their man - Derik Flint. Agent Flint (James Coburn) and at least a dozen buxom 20-somethings, scantily clad in current, mid-60s fashion, would save the world. The rest is history.
Besides Agent Flint's ample assistants, this spy-fi experience stuck w/me as a pioneering use of computing for talent recruitment, selection, and placement. That Agent Flint was retired from Z.O.W.I.E., yet still in the talent 'system,' was also interesting. The Z.O.W.I.E. brain trust could not identify the best talent for the job, but the trusty IBM Type 75, Model 2, Card Sorter could! (No mention of the backend, the actual computer, BTW. It was likely an IBM 650 Magnetic Drum Data-Processing Machine.)
To most people, the basic function of 'automation' or 'management systems' in HR is banal record-keeping. For example, SFA is for the sales manager to keep track of the salesforce. Salespeople despise it because their quota is what matters, and they are allergic to recordkeeping. Same for training/learning 'management systems' -- a recordkeeping appliance for the training administrator. So forth and so on for the entire HRIS GIGO hellscape.
For five years my responsibility was leading management succession for a large, diverse international firm. It included management succession technology. From day one the priority was the completeness, quality, and timeliness of recordkeeping on our stable of 150 top executives. Succession management is an important plank in talent strategy. An executive dashboard was furnished for the executives. They managed their own recordkeeping. (Trust me, when it involves succession planning, executives do a yeoman’s job at recordkeeping.) The succession planning system was a rare and stunning success in the HRIS portfolio b/c of impeccable integrity, and that senior management held responsibility for the complex process of selection and placement.
Resistance and rejection of novel job aids, process innovation, automation, machine learning, A.I., and so forth are often because the HR professionals know the record-keeping sucks and data integrity is poor. Once that is solved the embrace of higher-order computing is elementary. There are sound methods to achieve this level of quality.
In 1966 Z.O.W.I.E.'s Agent Talent System was protoscience. With the stunning wave after wave of innovations in automation, machine learning, and A.I., confidence is building for them in HR, and soon it will be routine to find and place Our Man Flint w/ease.
See: http://www.colabria.com/the-fiction-of-big-science/