In a study of 516 respondents, Capterra reveals how receptive Australian millennials are to the involvement of artificial intelligence (AI) technology in recruitment.
Key highlights from the survey include:
- 78% believe AI could help to reduce unconscious bias.
- 65% of Australian millennials think AI could make recruitment fairer in general.
- Most millennials want a hybrid of AI and human recruiters involved in the hiring process.
AI is changing the game for recruiters
Employing the right talent is critical for business success. Finding the talent, however, isn’t always simple—and hiring the wrong candidate is a costly expense.
Fortunately for businesses, some innovative AI companies are tackling this challenge head-on in a bid to improve talent acquisition. Here are four ways the recruitment industry is using AI technology:
1. More accurate advertising
AI technology can leverage existing data to better-focus advertising campaigns (i.e. showing them to the right people at the right time). Texio, for example, allows companies to tailor job postings to their target candidates by identifying the wording that will resonate best.
2. Speedy screening
Recruitment and HR professionals are increasingly using AI to make the first round of candidate cuts. The rules for eliminating a CV from the process is often based on the success of previous or similar applicants.
With this technology, recruiters can reduce the effort and time spent on this stage. According to a survey by LinkedIn, 67% of hiring managers and recruiters said AI saved them time.
3. Less operational burdens
AI sets out to automate low-level tasks on behalf of human recruiters. For Australian recruitment expert, Greg Savage, the involvement of AI will enable recruiters to be better sellers. In his article, he says:
‘Artificial Intelligence frees you up to complete the tasks which only you can do best: Influencing, persuading, advising, consulting, negotiating, collaborating, acting as an agent for the best talent, and building reputation and brand.’
An example of a company using AI to reduce operational burdens for clients is Filtered. They use technology to free up time for recruiters sourcing high-quality technical candidates by auto-generating coding challenges. While these won’t make up the entire recruitment cycle, it means recruiters can proceed candidates with greater confidence.
4. Automated feedback
Finding the time to prioritise feedback can be tricky. One company that is tackling this problem is Predictive Hire. CEO, Barb Hyman, reveals how her clients are using AI to focus on the entirety of the candidate experience:
‘Immediacy and convenience is the new norm for consumers, and candidates now expect the same when they are job hunting. Compare recruitment to applying for a bank loan where AI has been in use for over a decade. Why shouldn’t a candidate expect to get a response about a role within 24 hours too? AI can automate candidate feedback in a way that is impossible without technology.’
Taking the time to offer feedback to unsuccessful candidates can assist them in their job search. Constructive advice can also help relieve some of the disappointment from a job they may have emotionally invested in too.
Issues of bias are a concern
Critics suggest that using AI in recruitment promotes issues of bias. Other areas of concern include a lack of accountability and transparency of the hiring process, and issues with the data that the machine learns from.
A good example of this was an issue reported by Reuters about Amazon. The global business created an algorithm which unintentionally favoured male applicants over females for some of its job openings.
The AI based its decisions on a decade of CVs submitted to the company, of which had mostly been men. This was a reflection of male dominance across the industry, but as a result, the AI taught itself that men were preferable candidates. It even trained itself to screen out resumes that included keywords such as ‘women’s’ (as in ‘women’s chess club.’)
As soon as Amazon became aware of this issue, they changed the programs. However, it’s an event that may be enough to deter companies from using AI in their own recruitment efforts.
Omer Molad, CEO and Co-Founder at Vervoe suggests recruitment agencies and professionals thinking about using AI should try and keep an open mind. He explains how to avoid issues such as these:
‘Machine learning models learn from experience. If the data set that feeds the models is biased, then the models will learn biased behaviors. It’s incumbent upon recruiters to seek explanations from technology providers about how their AI uses data. It’s also important to remember that machine learning and other AI-based models need to be constantly monitored to ensure quality control. There is no “set and forget” with respect to AI.’
Looking at the candidate’s perceptive
Millennials make up 40% of the Australian workforce, according to Deloitte. As AI technology works its way further into the recruitment process, we wondered how receptive this generation will be.
We surveyed 515 Australian millennials to ask how they’d feel about employers using AI in the hiring process for a job they’re applying to. We set out to discover where the boundaries lie, from the initial application stage through to interviewing and skills testing.
65% of millennials say AI could make recruitment fairer
The majority of respondents said they believed AI could make the hiring process fairer. However, more than a third of respondents didn’t.
To explore the motivations behind these results, Capterrra asked respondents to explain their reasons.
Interestingly, respondents tended to argue the same point for and against AI technology. For example, respondents gave ‘an inability to feel’ as an argument to support both negative and positive feelings. Some felt this made the process more objective, as machines base their decisions purely on hard data. Others felt the lack of emotional intelligence and the inability to consider individual circumstances could make the process weaker.
Most millennials also believe AI could help tackle bias
The media often covers the existence of bias in artificial intelligence, and the majority of millennials have taken note. Just 19% of respondents firmly believe machines aren’t biased.
Despite this, 78% of millennials think AI could help to reduce bias when it comes to the recruitment process. 55% of this number said it could help to reduce bias, but only in some situations.
The results indicate that millennials still see the value in using AI technology in recruiting. Mostly, it’s because they believe humans always have the potential for bias (consciously or unconsciously), and with the right training, machines may be able to eliminate its presence faster.
One respondent said: ‘I think, provided biases are removed in the creation of AI (by using diverse developers and test subjects and research) this could have the potential of tapping into true equality.’
Millennials prefer a combination of AI and human recruiters throughout the hiring process
Many hiring managers use AI to screen out candidates during the initial stages of the hiring process—such as for selecting CVs that match their recruitment criteria. This is otherwise known as AI screening.
The rule-based system is the simplest form of artificial intelligence. It works from ‘if this, then that’ rules and mathematical formulas. Once a human recruiter creates a set of deciding factors in selecting who can progress to the next stage, an AI can automate it.
What motivates employers to use this method? Firstly, it speeds up processes not only for the recruiter but also the job seeker (successful or not.) Using an AI means screened out candidates are automatically informed so that busy recruiters can dedicate time to building relationships with high potential candidates.
Capterra discovered that 78% of millennials would be comfortable with a company using AI technology for this purpose. However, 61% of this number still want a human to review and check the CVs too.
The verdict: Companies could drastically speed up candidate sourcing using AI for pre-screening
Employers should take note that nearly a quarter (23%) of respondents said they wouldn’t be comfortable with this process.
Whilst there is a convincing argument to use AI in the pre-selection stage, businesses should offer complete transparency about its involvement. To avoid missing out on talent, recruiters could offer an alternative application route for those who only want a human assessment.
Using AI to check a candidate’s social media history
Another form of AI in recruitment is the analysis of candidates’ online activities on social media.
For example, DeepSense is a company that carries out assessments on candidate suitability based on an analysis of their online activities. For many, the idea of a potential employer reading our decade-old Facebook statuses is enough to make us cringe. However, AI tools are more sophisticated than that.
They look at learning ability, stability, attitude, autonomy, teamwork, and more. The software then assigns a personality score based on DISC profiles and the Big Five (OCEAN) personality analysis. These scores take into consideration the characteristics of successful candidates from the past to arrive at best-fit recommendations.
We asked respondents how they’d feel about an AI carrying out this task and 38% of candidates said they would be against it. More than a quarter (26%) would be comfortable for an AI to handle this task and 37% would be happy for an AI to analyse their social profiles, as long as a human also reviewed them.
The verdict: Candidates value human involvement for analysing social history
Recruiters could gain valuable insights from tools that provide this service. However, candidates still prefer human involvement.
While the numbers don’t heavily lean in one direction, recruiters should apply this type of technology with careful consideration. For example, the older the millennial, the more uncomfortable they feel about the notion. 40% of millennials aged 30 years old or more said they wouldn’t want an AI looking at their social history, compared to 32% of people under 30.
One way to approach this is to ask candidates to authorise the AI-powered assessment before it carries out the test.
The purpose of a job interview is to understand how a candidate could fit within a company. It’s also to assess whether their qualifications and career ambitions align with the position.
Sometimes, job interviews can be quick, but often the entire cycle involves the candidate meeting several members of staff over a few days. To speed up this process, many businesses are turning to AI.
Machine Learning (ML) technology is a type of artificial intelligence that recruiters increasingly use for this stage. In contrast to a rule-based system, the rules and formulas that the machine learning system applies are not set by humans. Instead, the system learns from patterns that lead to successful candidates in real-time.
One example where the recruitment industry is leveraging machine learning is the interview stage.
We wondered how comfortable candidates would be with a chatbot asking them basic interview questions. Most millennials were happy for companies to apply AI at this stage. 81% said they’d be comfortable either with or without a human recruiters involvement. Just 18% wouldn’t be uncomfortable.
We wanted to know how millennials would feel with AI playing a bigger role in their interview. For example, by conducting their entire interview.
There was a small increase in candidates feeling uncomfortable, jumping from 18% to 33%. However, half of the Australian millennials would still be okay with an AI application conducting their job interview, as long as a human recruiter reviews the responses too.
Body language, facial expression & tone of voice analysis
AI applications can analysis body language, facial expression, tone of voice and language at any stage of the interviewing phase. For example, L’Oréal and Coca Cola use software such as Seedlink to analyse the candidate’s use of wording in interviews.
The aim of this software is to gage a score of how successful candidates could be in the role. But would candidates feel comfortable being analysed in this way?
68% of respondents said they would. Of this number, 39% would want a human to also make an assessment. Almost a third of candidates (32%) said they wouldn’t be comfortable with companies using AI for this purpose.
Candidates expect companies to judge them fairly based on their ability. The results indicate that they may not trust methods like facial recognition to accurately determine their potential in a role. However, most candidates feel happier with a combination of AI and human analysis.
The verdict: AI is a good fit for first-stage interviewing (and maybe more)
Even though the above recruitment techniques are already happening in companies, millennials become warier the more involved an AI becomes.
With careful education, businesses should consider how AI can help enable recruiters during the interview stages of the hiring process. However, based on these results, it’s clear there is still a balance to be met.
Companies, such as Unilever are using AI and machine-learning to hire candidates based purely on merit. The second stage of the company’s recruitment process, after CV submission, involves a neuroscience game.
The game tests traits like focus, memory, emotional intelligence, and how risk-averse the candidate is.
The Dutch-British company said: ‘There is no right or wrong in the spectrum. Traits at either end of the spectrum could be really well-suited to different careers.’ As a result of using this technology, Unilever saw the ‘most diverse class to date.’ They’ve also experienced:
- Significantly more applications
- An increase in offers made to candidates
- A greater number of job acceptances
- A decrease in the average time taken to hire a candidate (from four months to four weeks.)
There are some obvious benefits to a company applying an AI-powered skills test to their hiring process. But how do candidates feel about being assessed this way?
The results indicated that the verdict is still out on using gamification to test candidate skills. A quarter of candidates oppose the idea while the rest would either like it or feel indifferent.
The verdict: Employers could do more to win candidates over
If employers are considering using games as part of their recruitment cycle, they should consider ways to bring candidates on board too. With so many respondents saying they’re indifferent, it’s possible that they simply need educating. As demonstrated by Unilever, using neuroscience games benefits the candidate too by increasing their chances of securing a job well-suited to them. In turn, this will lead to greater job satisfaction.
AI won’t replace human recruiters
Will AI replace hiring managers? It’s unlikely. But it’s probable that AI will integrate more and more into the hiring process to enable recruiters to operate smarter and more efficiently.
Prior to COVID-19, the recruitment industry generated $11 billion in revenue each year for the Australian economy. As the outbreak has forced businesses across all sectors to be more dynamic and innovative, AI-powered technology is likely to be a strong ally in the country’s economic road to recovery.
It’s worth businesses and recruitment agencies using this time wisely. Investigate the tools you’ll need to handle applications intelligently when companies begin hiring again. But more importantly, prepare ways to educate talent on how AI could play a bigger role in their future applications. A key driver of success for using AI within the recruitment industry will be Australia’s readiness for it.
Looking for recruitment software? Check out our catalogue.
To understand the opinions of millennials in Australia with regards to artificial intelligence in recruitment, we conducted an online survey between 3rd March – 8th March.
We surveyed people living in Australia from the generational group (aged between 24 to 39 in 2020) who work full-time, part-time, or are actively job-seeking. To mitigate the potential for bias, we screened out survey participants that worked within HR and recruitment. This left us with our final number of respondents for the survey: 516.