The toughest part of talent acquisition is screening the ideal candidate because, typically, it’s a pool of high-volume, low-quality applicants. As a result, potentially interesting candidates end up getting buried. But that’s not the only thing that stymies talent discovery. Candidates who do make it through the recruitment funnel often change their minds.
This is where AI in recruitment comes in. It makes the talent acquisition process faster, less laborious, and more efficient. So, recruiters have to do less with more and better outcomes.
Understanding AI in Recruitment
Artificial Intelligence (AI) is an umbrella term. Broadly, there are three types:
- Descriptive AI offers information about what’s happening.
- Predictive AI provides an insight into the future.
- Prescriptive AI tells us what to do based on the findings of the above two.
AI in recruitment is largely limited to the descriptive stage, with some aspects of prescriptive thrown in. A typical example would be automating manual tasks like reading hundreds of resumes, so an HR person is not bogged down by repetitive, high-volume tasks.
However, it is not the only application of the technology in hiring. From advertising to onboarding, AI solves problems throughout the recruitment process by streamlining the workflow.
Innovations in AI for recruiting
Innovations in AI in the past few years have completely altered the recruitment space. For instance, machine learning-based screening software can match and shortlist the best-fit candidates, while others can predict job performance.
One example would be an AI to rank candidates. They browse through applications and resumes to assess which candidates the People’s team should contact first.
Then some systems go beyond automatic relevant keyword search. Such AI compares the experience between resumes and prioritises applicants from the most pertinent to the opening to the least relevant.
Augmented intelligence takes this further by finding candidates through data collated from an internal database, social media networks, open-web profiles and job boards.
Digitised interviews are another vertical in intelligent hiring. The technology utilises video interviews to analyse facial expressions and assess personality traits in candidates.
It also grades verbal responses to predict job performance and how likely the applicant would thrive if hired. HireVue, used by Unilever, Vodafone, and Kraft Heinz, is one example of AI-led recruitment interviews. The technology has become particularly beneficial since the pandemic made virtual interviews par for the course.
Each technology has its place and its pros and cons in automating hiring. While an AI video interview, which allows for more elaborate and contextual responses, would better suit a behavioural assessment round, a recruitment chatbot, which assesses limited words, would better fit scheduling interviews.
Benefits of Relying on AI for Recruitment
Finding the perfect talent for an organisation requires nuances that only human experience can bring. Nevertheless, there are parts of the recruitment process that the manual system cannot handle as well as an AI can.
Think: searching for candidates and then reviewing resumes. AI can help take on such tasks. Moreover, it leaves the team free to use their strategic and creative abilities elsewhere.
Companies across the globe are already reporting positive results from automating recruitment that range from saved time, saved costs, and increased diversity. Here are the most prevalent ones:
- Correct candidate sourcing
- Saving time on pre-selection tasks
- Improving the quality of hire through automation
- Measure success chances via predictive analytics
- Shortlisting candidates through personality assessment.
Now let’s dig deep into each of these benefits.
1. Correct candidate sourcing
The most straightforward advantage of using AI in recruitment is targeting and sourcing suitable candidates. It has become possible because technology allows an organisation to target ads using historical online activity data.
Artificial Intelligence tools assess the browsing history of potential candidates and then show job vacancies to the right people at the right time. It’s an elementary application of AI but a valuable one.
2. Saving time on pre-selection tasks
Recruiters are extremely busy people, and a substantial chunk of that time is spent on menial chores like entering data into an applicant tracking system or sifting through applications.
However, by automating such duties, they can save time and better utilise it to engage with candidates and build stronger relationships. AI-driven chatbots are one way to do so. They understand natural language and can help remove manual processes like:
- Reaching out to old candidates for new positions.
- Answering simple questions.
- Screening applicants for the basic requirements of the position.
- Scheduling subsequent interviews by checking availability.
These are little tasks in the job selection process, but each needs to be done perfectly. AI can easily automate them instead of a recruiter wasting time on phone interviews.
Sense, Xor, and Celential are examples of tech companies creating chat bots for recruitment that solve different purposes. Some proactively reach out to candidates, others simply respond to questions when recruiters are not available.
3. Improving the quality of hire through automation
When the talent pool is large, sieving the right candidates becomes the most challenging part of hiring. It becomes even more complicated when organisations expand the scope to include passive candidates. That pool is more significant than people actively seeking a job change, exponentially so.
In such instances, AI can be capitalised to effortlessly dig through a deeper and broader pool in a shorter time. By sifting through the entire universe of potential applicants, passive and active, the search becomes inclusive.
It is also efficacious in ensuring shortlisted candidates meet essential criteria. Thus, improving the quality of hire. Moreover, as AI is used more and more, it gathers more data and learns what the recruiters are looking for, further fine-tuning the hiring quality.
4. Measure success chances via predictive analytics
A significant benefit of leveraging AI in job selection is data-driven decisions. Hiring managers don’t have to make gut calls on whether a candidate will succeed in the role or not. AI can use predictive analytics to measure the probability of how well an applicant will perform if hired.
Case in point: tools like Pymetrics look for traits and signs in new candidates that previous hires had. To make the AI more precise, they gather data from not just all candidates that were interviewed or hired but also those who succeeded in the position.
Essentially, the AI looks at data from the entire life cycle and does so in seconds to deliver valuable information and insights. The same job would be extremely difficult for a human. A data analyst or scientist could do it, but that would take hours.
5. Shortlisting candidates through personality assessment
One last but certainly not the least benefit of artificial intelligence in recruitment is that it assesses talent’s personality by considering information that human recruiters cannot, at least not as quickly.
AI is gamifying job selection to judge cognitive and intangible traits and assess if the applicant will be a good fit for the open position.
One example is checking impulsivity, attention space, and how they learn from mistakes in applicants by counting the number of times a candidate presses the space bar when a red circle flashes on a computer screen.
The AI compares the applicant’s data with top employees in the role they have applied to deliver results.
Challenges of Applying AI in Job Selection
There is no doubt that in some stages of recruitment, AI has the edge over humans. However, it can be argued that the result is not guaranteed to be accurate and lacks accountability.
But more importantly, AI can learn human biases, with assumptions being built into the algorithm due to erroneous data.
The necessity of volumes of data
AI is trained through data, volumes of it. For example, in the case of recruitment, it uses historical data like the resumes of previous or similar applicants coupled with the actions of hiring managers.
It’s how AI learns to mimic human behaviour. It is the necessity of mountains of data to make accurate predictions that becomes the first hurdle for AI.
Once the algorithm is deployed, the corpus of data keeps growing. If it’s not monitored consistently, it gives erroneous results instead of accurate ones.
The reason for it is “model drift in machine learning,” which is the tendency of the algorithm to go off course and away from the desired outcome. The only solution is constant human scrutiny of AI.
The risk of a biased hiring process
Because AI is trained through data and to identify patterns based on previous behaviours, it can become biased. There is always a risk of the algorithm inheriting both the subconscious and conscious preferences of recruiters.
Two real examples of AI bias are Amazon and LinkedIn.
Both created AI to recruit candidates, and in both cases, the algorithm picked up more male candidates over females. To counteract such bias and ensure qualified and stellar candidates make it through the recruiting funnel, AI necessitates perpetual surveillance.
“The real safety question, if you want to call it that, is that if we give these systems biased data, they will be biased .”
Another challenge with using AI for hiring is the hesitancy of recruiters. Not all people are ready to embrace technology because it is easy to “hack” it.
For instance, many software looks for keywords like lead generation, action, project management, and growth in resumes. A candidate with access to such insights can skew the AI towards their CV by simply including these keywords even when they don’t have the necessary experience.
That said, the alternative is not any better. Hiring managers would need to carefully read and evaluate every resume, which could take weeks, if not more when the number of applicants is massive.
People and AI Work in Tandem in Recruitment
It is not an “either-or” question. Instead, it’s a partnership of technology and human skills. Where the first is good for economies of scale, the latter is excellent for nuances.
An AI-powered video interview may catch subtleties in applicant behaviour that a human being may not. But is the AI sensitive enough to know that nervousness is the root cause of the behaviour? It’s context like this, and empathy that technology lacks. At least for now.
On the other hand, recruiters are constantly motivated to shorten their hiring time and improve the quality of their hires. But it’s impossible to get through thousands of applications quickly while being thorough. AI tools solve this problem of efficiency and effectiveness. They can swiftly screen through huge applicant pools and do it cost-effectively.
It’s only when AI and HR teams are used in tandem that they can completely re-imagine the recruitment process and bring it into the 21st century.
Technology can take over ordinary processes that no one appreciates, like making the first round of cuts, while people can focus on discovering better ways to filter the best of the talent.