Could a machine choose your next hire?
Human Resource Management

Could a machine choose your next hire?

In Chile, one job recruitment company, AIRA, is using artificial intelligence to help make hiring decisions based on emotion analytics. The company has created a system that publishes vacancy announcements in the most widely used recruitment websites. It then reads and ranks all résumés, applies psychometric tests and conducts video interviews with applicants, and assesses the applicants’ performance with emotion analytics, which translates their attention levels and facial expressions into numbers. "At the end of this short process, human recruiters can focus their scarce time on conducting in-depth interviews with the best qualified candidate,” according to an Accenture report. Are machines becoming the deciding factor in new hires?

The simple answer is yes. The use of machine-driven recruitment processes is increasing as artificial intelligence evolves and becomes more sophisticated. There are already companies around the world looking at how AI and other technologies can not only make the hiring process easier but also ensure a better fit between candidate and company.


How is AI changing the hiring process?

London-based AI recruitment platform Elevate Direct, for example, uses cutting-edge machine learning techniques to provide decision support for recruiters and hiring managers. The company provides a ranked list of relevant candidates for a given job, enhancing the productivity of the recruitment process by up to 2.5x, according to CEO Dan Collier.  

Realistically, artificial intelligence is not actually all that intelligent - at least, not yet. Artificial lack of intelligence: how smart is AI really?

California-based Mya, which was launched just nine months ago, is helping major recruitment agencies and Fortune 500 companies overcome hiring challenges using AI, proprietary machine learning and natural language processing to automate recruiting from resume to hire. This includes sourcing, screening, scheduling, on-boarding, and engagement.

Eyal Grayesky CEO and co-founder of Mya Systems, the creator of AI recruiter, Mya, says it is a great example of how AI offers a better experience for applicant and recruiter and provides useful data to help ensure hires are a good match to the company. In fact, a study of 15,000 applicants that Mya processed for a large customer in sports manufacturing found that Mya reduced time-to-hire by 70%, increased candidate engagement over 150% and that candidates were three times more likely to hear back from the recruiter if they responded to Mya’s questions.

Collier admits there have been hurdles in coming up with the right product for a given market, as some AI technology behaves in ways that are counter-intuitive and it is a challenge to make it explainable. “For example, an earlier version of our product had counter-intuitive rankings, and this really hurt the customer experience,” he says. But Collier is optimistic about the potential of AI to be a useful recruitment tool, particularly in combination with human recruitment agents.

Tom Quayle, head of workforce analytics at The Chemistry Group, emphasises that machines are already choosing companies’ next hire. “Three-quarters of applicants are screened out of a recruitment process based on the words on their résumé, and tools such as psychometrics, video and emotion analytics are being deployed to further reduce the remaining 25%, such that every five in 1,000 applications end up in front of a hiring manager,” he says.

The possibilities for how AI and automation is integrated into recruitment are varied. The Chemistry Group worked with SAP to develop a recruitment game using Facebook as the platform, something the enterprise application software development company had been wanting to do for years.

The game The Chemistry Group has developed for SAP not only creates a more engaging experience, but it also allows them to observe how players interact and play the game. SAP captures this data and correlates it with bespoke assessments that Chemistry has deployed across SAP, further down the recruitment funnel. When players log in with their Facebook accounts, data on their ‘likes’ creates a personality profile, using an API designed at Cambridge University, a partner of The Chemistry Group's 'Lab’. This data is then correlated with their game behaviour and the data from the Chemistry assessments.

“Combining these data sources gives us a rich insight into the behaviours, motivations and values of the individual players. This allows SAP to match individuals much more accurately to their unique WGLL way more accurately than they would be able to using traditional assessments or recruiting methods,” Quayle says.

Over time, SAP will be able to model that behaviour against the personalities of their own high-performers, so they can predict which players would be successful at SAP and offer them the opportunity to apply. “This kind of technology allows SAP to move away from laborious, formal psychometric testing and is far more engaging for a potential employee. It also allows them to attract a much wider talent pool of people who may not have considered directly applying to SAP,” he says.


Is AI a replacement for human skills?

Despite interest in the ways in which these technologies can make recruitment more efficient, almost all of those involved in the field agree that AI should be a complement to human recruitment.

Chris Nicholson, CEO of Skymind, says that AI should serve as decision support, an automated second opinion that offers recruiters an additional perspective, and may lead them to double-check their decisions.

“When there's simply too much data for humans to review by hand, then AI can help surface the most interesting bits. In this case, it could help surface job candidates that might otherwise be overlooked,” he says, but adds that when humans have expertise in a field, they shouldn't be excluded from a decision making process.

“They should work with AI-based filters to come up with the best decision. And that will lead to better decisions in many cases, because humans and AI algorithms have different strengths and weaknesses. Humans understand things about the world that computers don't. On the other hand, algorithms don't get tired [and therefore] they don't make more mistakes in the late afternoon, as humans are prone to do.”

Martin Ewings, director of regional sales and specialist markets at Experis, agrees that where AI is really powerful is where it is harnessed as augmented intelligence rather than artificial intelligence. “At its most effective, this technology is used to augment the human role and not replace it. Recruitment will never be an exact science and as a result, the human aspect of hiring will always be vital,” he says. 


What are the limitations of AI recruitment?

Carrie Graham, international executive search consultant at MRL Group, says that AI is not a good fit for all types of recruitment, stressing that recruitment is all about relationship building with both candidates and the companies that are recruited for. “I believe you need to speak to the person about their career aspirations, culture, team size and where they would see themselves in their next role,” she says. “If we use AI to screen candidates you would not be able to ask these questions.”

Graham adds: “While I believe AI could support some recruitment consultants, I don’t believe it would be suitable for the highly skilled tech market I recruit within where the skills are finite and there is a real need to speak to each and every candidate to ensure they are a right fit for our client base. Some of the best candidates I have found have not had all their details specified on their CV and it has taken my discussions with them to establish that they have the right skills and attitude required for the position.”

Peter Linas, international managing director at Bullhorn, believes that recruiters should not feel threatened by automation and what AI has to offer. He points out that recruiters have always been early and enthusiastic adopters of technology, using job boards not long after the internet was launched and leveraging LinkedIn to secure talent from the moment the platform was launched.

“Recruitment is fundamentally a people-facing profession: it’s about making connections and turning them into relationships – about matching the right individual to the right organisation. A machine can’t form an authentic connection with a human, and barring any major technological leaps, it won’t ever be able to. If recruiters fear obsolescence, they shouldn’t,” he says.

Linas admits, though, that automation is going to take on some of their job duties, but says these will typically be ‘low value’, yet important tasks, such as identifying candidates, contacting them via email, figuring out mutually appropriate meeting times and places, and most forms of data entry.

“In this respect, an automated task bot acts as personal assistant to – rather than a replacement for – the recruiter. It takes on work that doesn’t require specialist intervention, freeing up time to focus on the bread and butter of recruitment: building trust and value in client and candidate relationships.”

Ewings says that AI and automation can be used to streamline and enhance the recruitment process, pulling in and analysing huge volumes of data on candidates, and highlighting those that need more time spent with the recruiter to secure them a position. The analytical intelligence can also draw on data and insight from a range of other roles, as well as the attributes of successful candidates in each, to suggest alternate avenues for struggling candidates.

He cautions, though, that the downsides of AI for the employer or recruiter mean that you lose some of your ability to ‘sell’ the opportunity to candidates. “If the candidate has a number of offers on the table, it becomes harder to differentiate your organisation from the competition and recruit the best talent,” he says. “This is where the human element really adds an invaluable extra layer.”

Ewings adds that, in its current format, AI solutions only reach active candidates in the open marketplace, meaning companies could potentially miss out on a great passive candidate who could be open to the right approach.

Grayesky suggests that businesses wanting to implement an AI recruiter should realise that AI is ideal for high volume recruiting needs, such as hiring the 30,000 warehouse workers that Amazon recently announced it will need. “It’s also ideal for call centre, retail, hospitality, and food service. It may not be the ideal fit to recruit for highly skilled positions, such as CEO, CFO, and so on,” he says.

They should also understand that AI can go beyond gathering basic information, such as shift scheduling, pay range, physical requirements and use sentiment analysis to gauge a candidate’s interest and motive. It may also be able suggest alternative jobs if a candidate is not interested or is not a fit for the first position they applied for.

Beverley Nicholas, regional talent director at PageGroup, says businesses wanting to use AI ad automation also need to understand what candidates expect when applying for a new role. “Though many professionals could be open to the idea of a smarter and more accurate system such as AI, they will still expect an easy and reliable process,” she says. “If they are spending more time filling in forms and going through an automated process over face-to-face interactions, this will affect the way that the candidate perceives the recruiter.”

For Nicholas, the main challenge is finding a way to implement systems that are reliable, transparent and able to identify the best talent for the most relevant role, while considering the applicants and ensuring that their user experience does not run the risk of driving them away or discouraging them from completing the application process. 


«Fintech is the latest niche to become hot for bespoke training


Africa needs data scientists: What will it take to train them?»
Bianca Wright

Bianca Wright is a UK-based freelance business and technology writer, who has written for publications in the UK, the US, Australia and South Africa. She holds an MPhil in science and technology journalism and a DPhil in Media Studies.

Most Recent Comments

Our Case Studies

IDG Connect delivers full creative solutions to meet all your demand generatlon needs. These cover the full scope of options, from customized content and lead delivery through to fully integrated campaigns.


Our Marketing Research

Our in-house analyst and editorial team create a range of insights for the global marketing community. These look at IT buying preferences, the latest soclal media trends and other zeitgeist topics.



Should the government regulate Artificial Intelligence?