As in many fields, artificial intelligence (AI) is playing an increasingly pivotal role in recruitment. AI-enabled tools are being deployed at every stage of the hiring process, from sourcing candidates and candidate engagement to interviews and onboarding.
The potential and desired outcomes go much further than simple automation of processes, with some incredibly clever recruitment tools being explored. But it will take some time before the full potential of true AI is realised.
Consulting firm Mercer’s 2019 Global Talent Trends report found that 88% of companies globally were already using AI in HR, with 44% identifying the best candidates based on publicly available data, 41% using chatbots to engage with candidates, and 40% screening and assessing candidates with the assistance of AI.
Richer talent pools
Locating the most suitable candidates is, of course, key, and recruitment professionals may struggle to adequately screen large numbers of incoming applications.
‘Using AI-enabled software for CV screening is particularly useful for large organisations that attract thousands of candidates for any given role and can bring significant economies of scale,’ says David Suarez, people and organisation advisory leader at PwC Middle East.
‘Employers can make meaning-based search suggestions that are far more accurate than text-based equivalents’
On the sourcing side, tools such as Entelo, HiringSolved and Textkernel greatly widen the talent pool by rapidly sifting through millions of unstructured data points, such as CVs on job boards, agency databases, LinkedIn profiles, social media and applicant tracking systems.
More advanced tools can perform semantic searches that judge what is meant or inferred, as opposed to simply what is written, in a CV or social media post. Middle East jobs portal Bayt.com, for example, has developed a meaning ontology for its CV search technology.
‘Employers can make meaning-based search suggestions that are far more accurate than text-based equivalents,’ says Bayt.com CFO Nauman Asif Mian FCCA.
With accurate data in massive quantities, such tools can unearth high-value ‘passive’ talent – candidates who may be open to a new role but are not actively seeking one, who make up 70% of the workforce. For example, a feature in Entelo’s software analyses variables, such as length of time in a current role and increased social media activity, to predict a candidates’ receptiveness to new opportunities.
Bias and diversity
For AI tools to be effective, data must be clean, accurate and free from bias. In 2018, Amazon scrapped an AI CV-rating tool after the system showed gender bias in favour of men for technology industry posts, due to its use of historical data that reflected the sector’s gender imbalance.
Platforms such as SeekOut and Entelo (used by PayPal, Target and Lyft) address bias and diversity in candidate sourcing by anonymising information on gender, length of career or ethnicity. They may also include filters to target under-represented groups; SeekOut, for example, deploys AI to analyse aspects of a profile and classify potential candidates according to diversity categories.
Better candidate experience
AI is being harnessed to improve the experience for candidates, too. PwC’s Future of Recruiting survey found that 49% of candidates had turned down an offer due to a poor recruiting experience.
Platforms such as HireVue, used by PwC, and Mya, used by Deloitte, Hays and L’Oréal, are at the forefront of automating and improving candidate engagement, using AI-enabled chatbots, automated text, social media and email engagement, interview scheduling and video interviewing.
At Hays, Mya is used to re-engage with potential candidates whom the recruiter knows but has lost touch with, leading to 41% of recipients responding with useful information.
‘That’s a great result – much better than other channels we’d tried, and certainly better than we anticipated, says Jacky Carter, Hays group’s digital engagement director.
‘Consultants are calling people who want to be called, candidates get to update their information at a time and in a format that’s convenient for them, and hiring managers get to select from a talent pool that is bang up to date.’
Meanwhile, Dubai-based platform Searchie uses image recognition software powered by machine learning to evaluate candidates’ facial expressions and determine how they will perform in a real-world setting. Technology also exists that analyses speech to assess personality traits and soft skills.
Where next?
It’s only logical that AI-based recruitment applications will expand further into more complex tasks, particularly in the predictive sphere, believes Suarez.
‘I expect a shift towards tools that try to optimise parameters such as quality of hire and likelihood of attrition,’ he says. ‘There are already several platforms out there that claim to do this, but it’s still early days.
‘More needs to be done around wider training and testing of data sets to reduce algorithmic bias before these applications can truly add value.’
Henry Morse, associate director at recruiters Robert Half UK, points out that despite the hype, AI isn’t a recruitment panacea.
‘For all the technological upgrades we make, recruitment remains a people-focused industry, where specialists take the time to understand the full range of what candidates and hiring managers are looking for,’ he says. ‘These functions work to help human consultants, not replace them.’
