Written by Maciej Szczerba

Will machine learning replace recruiters?

As with many other professional professions, the recruitment profession appears to be under serious threat from the development of machine learning. Will recruiters be replaced by chat-bots? Will sourcers / researchers be replaced by a self-learning search engine?

As with many other professional professions, the recruitment profession appears to be under serious threat from the development of machine learning. Will recruiters be replaced by chat-bots? Will sourcers / researchers be replaced by a self-learning search engine?

I will start by saying that, having worked in the recruitment industry for more than 12 years, I have heard the prophecies of the 'end of the world of recruiters' several times. The prophecies were centred around three main themes:

Prophecy 1 
Social networks will allow candidates and hiring managers to communicate directly, without the participation of recruiters - intermediaries (whether external-agency or internal-employed in HR departments).

Prophecy 2 
Recruitment advertisements will simply disappear if the "hiring managers" can find the candidates directly online themselves.

Prophecy 3 
Employer branding and recruitment marketing will replace the recruitment service. Like inbound marketing / inbound sales, image campaigns and a variety of advertising activities aimed at external candidates and ensuring a good reputation as an employer among internal employees were supposed to make these activities respond to those people who companies want to recruit to their ranks.

My start in recruitment in 2008 was when social networks broke out, including 'professional contacts' such as LinkedIn and Goldenline. I saw the moment of growth of both of these networks, the moment of GL being pushed out by IN, and the moment of subsequent stagnation of Linkedin.

I watched 4-5 years ago a dump of various kinds of "super-candidate search engines" searching GitHub, StackOverflow and other social networking platforms, which were supposed to revolutionize the market of searching for candidates in the IT industry and then instantly conquer the market of candidates in the medical industry (the second industry globally suffering from a structural shortage of candidates).

In the meantime, I have heard many times that literally "in a moment" advertising portals (such as pracuj.pl) will disappear like the famous Monday paper supplement to "Gazeta Wyborcza" from the 90s.

To paraphrase the song "Electric Guitars" - and what... and nothing...

Well, maybe almost nothing...

However, in 2020, in the light of two huge economic and social events such as the Covid-19 pandemic and the significant increase in 'cloud' computing capacity in recent years, I would not dare to be arrogant and claim that these phenomena will not affect the market for recruitment services. 

They will have influence.

Yes, I believe that machine learninging/AI will change the market for recruitment services (internal and external) to a huge extent. The profession of sourcer/researcher can be seriously threatened. In the end, however, the profession of recruiter, whose history dates back to the Roman Empire, some 2 000 years with a poultice, will not disappear from the face of the earth. It will be automated, it may lose weight, but it will not disappear.

Whatever you say, nuances are important in recruitment. For example, in our field of IT recruitment, we sometimes get negative feedback on the specialist presented: "too little technical skills". Feedback presented in this way is practically useless, because what does it really mean? Using the example of the position of a Java programmer, and thus completely randomly, let me quote the default questions that I am currently asking myself:

- Did the candidate not know the latest version of Java?

- Did the candidate have problems with the creative solution of the problem (if a programming task had to be solved during the recruitment process)?

- Is the candidate not familiar with microservice architecture solutions?

- Does the candidate not know about cloud solutions?

And so on and so on...

I wanted to cite this example to show how difficult it is to detail the phrase 'technically unfit'. We, experienced recruiters in all the industries for which we work (not just IT), know that often the job description for which we are looking for the right employee does not reflect the actual needs and expectations of a hiring manager. Therefore, we always start our activities with a detailed interview with a hiring manager. A good recruiter also reads a lot "between the lines" from such an interview.

For the time being, we are a long way from the moment when the AI algorithms are able to sense these nuances. Machine learning algorithms must have well-prepared data to see patterns in them. Contrary to appearances, the CV set is much more diverse than, for example, a set of medical photos, a set of financial data, or data on product sales. Each person puts together his or her CV in a different way, emphasizing his or her strengths in a different way.

In my opinion, the AI algorithms are not at this stage of cognitive learning to assess all these details.

Contrary to appearances, sourcer's work is not only repetitive - it is also largely creative, even strategic work. In the United States, the profession of major sourcers is often carried out by very experienced people at the peak of their careers. They are characterized by, among others, excellent knowledge of Boolean logic, often the ability to read the web code (yes, yes!) and excellent communication skills. Taking into account these arguments and statements made by AI experts, I think that we are a decade away from arranging search strategies through algorithms. 

However, there is certainly a growing interest in supporting AI in its work as a recruiter. It is worth quoting here such tools as German honeypot.io (for the IT industry), which are a hybrid of recruitment agency and application. The American hired.com, which was probably the first solution of this kind, is also something similar. Similar tools /agencies are constantly appearing - recently I came across such a solution for the American sales & marketing industry.

When I talk to clients about such solutions, I find that people, however, buy services and goods from other people and expect that, on the part of the recruitment service provider, a person (not a machine) will talk to the candidate before his or her recommendation. Even if a person is assisted by AI (which, I predict, will become a standard within a few nearest ones).

What is your opinion? In which areas can a machine replace humans?

Previous Story
Where is Poland on the world map of IT and outsourcing part 2?
Next Story
HR and Recruiting predictions for 2021

Let’s stay in touch! Subscribe to our free newsletter

To be notified of new blog posts, webinars and training videos on recruitment and hiring