Benefits of AI in Talent Acquisition
A recruitment partner based in Shanghai told me that they now typically receive 260 (new) applications for an open position in a large company. The hospitality sector garners more than others: in Manchester, UK almost 1,000 applied for a receptionist job in a single day in July 2020!
You can understand why there is rising excitement from recruiters looking to use computer technologies and Artificial Intelligence to sift through those applications. Feed the AI 100 resumes and out pop the top 5 candidates. Or at least that’s the utopian ideal. [todayonline.com].
In theory, an AI could reduce the time to sift through applications and, without bias, identify the candidates with the right set of attributes and skills and experience and qualifications that the position requires. These candidates could be tested remotely using psychometric tools, games, activities, or simulations that demonstrate their fit for the role. And video interviews of candidates answering a series of questions could be parsed, transcribed and analysed for spoken content, nuance, body language. But will it accurately get the meaning behind the words used?
Automated bots, machine learning or full-blown AI, the promise of time-saving, unbiased, more accurate recruitment is alluring for many companies. Would you use an AI to recruit talent for your team?
Will your talent pass the AI test?
What’s it like to be rejected by a computer then?
Several mid-career professionals, most of whom have been retrenched thanks to COVID, others at the end of a contract or project or business collapse. All experienced, well qualified and willing to work, are finding it increasingly challenging to find suitable employment. Most have at least 100 applications “out there” and an average of 4-5% hit rate for getting an interview. Most of those get tested using a psychometric or one of the growing number of AI personality or skill tests.
Then they hear crickets. Nothing. Nada. Rien. Zip.
A few have been on several rounds of interviews. Then… silence.
A month or two later they rediscover their letter box and the letter we all love to receive: “Dear Mr Smith, We’re sorry… We’ll keep your details on file…”
No explanation. No suggestion as to why you weren’t chosen. And if you ask and get a response, it comes on the lines of: “… a better candidate…”
Yes but what was wrong with me?
Enough of these and even the most zen have been known to crack. With no honest feedback, no test results, no guidance towards other, more suitable roles available.
Of course, there is the occasional celebration.
“We are pleased to offer…”
Interestingly, most candidates don’t ask “what was right with me??
Rejection is tough to bear. Would you prefer to be rejected by a human or a computer?
“Judge not, that you be not judged.” NKJV Mat 7:1
And HR is already behind the curve when it comes to the real-world application of AI. It is possible that AI could have a transformative impact on talent attraction, retention, and development according to McKinsey. And the biggest value an AI-enabled HR team can provide to a business is improving employee performance.
Recruiter or Candidate?
I have a couple of questions for you at this point:
- Would you use an AI system for recruitment for your team?
- Would you like to be a candidate being assessed by an AI?
Trouble in AI Paradise
The system is flawed now. Add AI to the mix and you get a flawed system run by an AI.
I added one word to my LinkedIn profile recently and suddenly had a deluge of “suitable” opportunities. I had stumbled upon one of the keywords that the LinkedIn boffins included in their Qualified Applicant Artificial Intelligence model for these roles and now, it seems, the combination of skills and experience match more vacancies.
So I reached out to a few LinkedIn Profile Makeover specialist to get their take on keyword stuffing on profiles (sorry, they prefer to use “optimization”). And yes, of course, you can get your profile made-over quickly and easily to better match the jobs and openings on offer. There's even a couple of AI powered software platforms that will do this automagically. But isn’t that trying to cheat the system? Well, let’s face it, many CVs are a fiction of fancy job titles, mega responsibilities and fantastical achievements.
If I know or suspect that a keyword is being used by the Applicant Tracking System then I make sure to include it on my profile.
(BTW If you're after a 2021 tutorial on this have a look at
Daniel Lorenzo over at Let's Eat grandma
tips to help your resume stand out in the age of AI from Today
But the promise of AI is that a computer isn’t biased! Surely that’s a good thing?
Online retail giant Amazon didn't agree. In 2018 it was widely reported to have scrapped its own system, because it showed bias against female applicants. The Reuters news agency said that Amazon's AI system had “taught itself that male candidates were preferable” because they more often had greater tech industry experience on their resume. Amazon declined to comment at the time.
Prof Sandra Wachter, a senior research fellow in AI at Oxford University concurs that ”So in recruitment, looking at the successful candidates of the past is the data you have. Who were the chief executives in the past, who were the Oxford professors in the past? The recruitment algorithms are going to pick out more men.
Amazon's computer models were trained to vet applicants by observing patterns in resumes submitted to the company over a 10-year period. Most came from men, [todayonline.com]
There are significant challenges for AI systems according to James Meachin of UK business psychology consultancy Pearn Kandola is a specialist on the recruitment sector. He says that AI systems still have a number of challenges. [bbc.com]
“The first step in selecting candidates is to correctly parse what they have said or written,” he says. “On this basic level, leading voice assistants from Google, Amazon and Apple still routinely fail to understand what people are saying.
“If an AI system can accurately transcribe what has been said, the second – greater – challenge is to detect the meaning embedded in those words, the semantics, nuance and context. Here, AI systems can fail to understand. In contrast, a human listening to the conversation will intuitively understand what is meant.”
In Europe, the potential for bias with AI has come under great scrutiny, particularly since GDPR implementation. “Bias in AI is well known.” [talentguard.com]
Talentguard go on to say that Bias is the Achilles heel of Talent Management solutions that employ AI in any form or fashion
For a robust research paper on this, check out Why fairness cannot be automated: Bridging the gap between EU non-discrimination law and AI.pdf
Fix the system and AI can add serious value
AI is not the panacea for many of their most challenging talent management problems. It's impossible for AI to fix the flawed system it is modelled upon.
When your LinkedIn profile is a true and accurate reflection of who you are, what skills you have and your experience (so, nothing like a typical CV then :-)) AND the recruiter has truthfully and accurately identified the skills and experience required of open positions, then recruitment with the help of AI could be a boon to both recruiter and recruited.
According to McKinsey, when you take humans out of the hiring process and you’re using fundamental aptitude versus tacit knowledge that we’re testing for, that actual ability is very evenly distributed across society. That's a promise that AI can help remove bias throughout the employee life, from recruitment through reward, performance, promotion, development and compensation.
For example: using an automated bot to better guide applicants to positions they are suitably qualified. The goal is to make sure that every candidate or employee is hired based on what they can do and their potential for learning as they pursue an aspirational role.
Once you apply and give access to your (honest) profile data, the AI then matches candidates who have those skills with open positions.
The biggest value an AI-enabled HR team can provide to a business is improving employee performance.
AI can enable HR to apply and scale learning and support in new, highly-personalised ways, well beyond traditional training courses. Though most organizations continue to perceive workplace learning as programmatic and focus on building and delivering courses.
Many of the more famous Talent Assessment companies now have an AI version of their psychometric tests. Some are true AI, many use machine learning, more are automated bots. The trend is there and the market huge. But if you eliminate candidates solely based on the results of a behavioural profile (especially with norms established 50 plus years ago on a largely Caucasian US Male base) – then your AI is going to be programmed to miss an awful lot of real talent too.
Can AI identify “Talent”
There is great potential for AI systems to identify true talent and the core values and attributes that make for your ideal team member.
Tired of seeing the abuse of DISC, MBTI, Big 5 and many others, years ago, we developed our GAPPS Assessment to help identify strengths and weaknesses of someone’s Attributes, Abilities and Agility. Unlike other tools that show how you measure up to the average American, GAPPS results are benchmarked against a subset of “successful” individuals in the same department or role, same cultural or country context and same organisational level. This enabled us to identify if someone is likely to be successful in a particular role, and if there were some gaps, how these might be bridged.
Many people disliked their results at first because, if they had seen previous test results, they usually see how magnificent they are. GAPPS is honest and shows how you compare with successful people in your field. Would you rather be benchmarked against an average person or one who is successful?
Critically, every person taking the GAPPS assessment receives feedback and coaching. This means that, when used in recruitment, a candidate gets feedback on their strengths, weaknesses and development priorities for a specific role. In the event they were less suitable, the system helped identify roles, cultural contexts and organisational level that may be a better fit for them.
Can an AI spot those words? Great coaches and interviewers have known this for centuries – they probe without judgment when needed to uncover hidden gems buried in a candidates unconscious memories and help them to find out who the real person is behind the protective mask.
We have found that talents and strengths, values and attributes can be uncovered in the stories we tell about ourselves, not in that work of fiction we call a CV. It’s the stories of significant moments in life from an early age to recent events. In our feedback coaching we help candidates uncover their stories which helps interpret their GAPPS scores in light of their experiences, their real talent and attributes. Put simply, we use discernment and wisdom gained through years of experience to help guide candidates on their best path that supports their prospective employer.
Even when a candidate is rejected for a role. They speak highly of the employer because the experience of their assessment and coaching was valuable to them. Some candidates went off to gain the experience and skills they lacked, others were recontacted when more appropriate roles becomes available. Team collaborated more readily because everyone on the team knew their strengths, weaknesses, and leveraged each others true talents.
Can an AI do all that? The only real stumbling block that I can see is that AI has no wisdom or discernment. Other than that, if we build on a good foundation of practices, processes and data, AI can help HR a great deal. If you build it on the current, largely flawed system evident in most organisations worldwide, you’ll have a flawed AI version of the system.
Want to learn how GAPPS can help you recruit the right talent, build collaboration in your team and improve the bottom line? It’s not a true AI, but it comes with something far superior: the wisdom and discernment of Human Intelligence.