Automation, algorithms, and machine learning are revolutionizing recruitment & hiring. How can recruitment professionals stand out from the crowd?
There was a time, not so long ago, when applying for a job was relatively straightforward. You had to print out a well-presented CV, dress smartly and prepare for a face-to-face interview.
But these rules no longer apply. Over the past two decades, digital technologies have radically transformed the employment landscape. Automated software, professional databases and one-click applications now dominate the hiring and recruitment process.
Automated recruitment is becoming increasingly widespread. ATS, or application tracking tools, are used by employers to sort CVs, analyze and schedule interviews. AI is also increasingly used for predictive analysis and performance evaluation. Today, it’s not uncommon for candidates to be rejected by an algorithm before even being contacted by a human recruiter.
The world of employment is set for a revolution with the introduction of artificial intelligence-based recruitment tools. Human resources managers find themselves in a difficult situation, with shrinking budgets and a growing influx of applicants, the result of the economic recession and the post-pandemic boom in teleworking. This automated software, which makes crucial hiring decisions without human oversight, raises important questions about privacy, accountability, and transparency.
From a jobseeker’s point of view, AI-based recruitment software is like a black box. You can spend a long time submitting an online application, only to receive a standard rejection email with no explanation. This dehumanised process can be disconcerting. ebadolahi, Senior Project Manager for Economic Justice at UPturn, a for-profit organisation focused on technology and equity, says: “No one really knows what happens when you navigate through this process”.
However, this technology can also bring considerable benefits. Various online tools, such as CV enhancement software that optimizes keyword matching and generative AI platforms that help write cover letters, make it easier for candidates to navigate the HR selection process. What’s more, thanks to algorithm-based professional networking platforms such as LinkedIn and Indeed, access to job vacancies is wider than ever.
When asked about the possibility of automation taking over the recruitment process entirely, most experts maintain that recruitment remains a people-driven process. Adapting your application to an applicant tracking system (ATS) is just one step in opening the door,” explains Ankur chaudhari, product manager at jobscan, an online tool that optimises CVs. He compares the process to an entrance exam: even if you get a high score, you still have to compete with other candidates for a place at a prestigious business school.
Jobseekers will always be key players in the hiring process, whether AI is used or not. Understanding the rules of the game doesn’t change this reality, but it does give you a head start.
Lauren Milligan, a career coach and CV-writing expert based in Illinois (USA), works with people who have taken a career break. These individuals, reluctant to submit to an artificial intelligence assessment, turn to her company, ResuMayday, for help.
For large companies that manage thousands of applications, automation offers a way of relieving administrative tasks, increasing efficiency and reducing costs. Nearly 99% of Fortune 500 companies use application management systems such as Worday, Taleo, Jobvite, Greenhouse or Lever to sort candidates.
Automated tools are versatile and can be used at different stages of the recruitment process, from assessing skills and personality to monitoring body language during interviews and even examining social media accounts.
Most employers purchase these tools from third-party platforms and customize them to meet specific recruitment needs. For example, Microsoft works with the Humanly platform to conduct automated conversations and virtual interviews with candidates. However, Humanly does not take on the entire recruitment process, says CEO Prem Kumar. Instead, it acts as an assistant to recruiters, helping them to take notes in the applicant tracking system (ATS), write emails and check candidates’ references.
The lack of transparency and supervision is at the heart of the problem. Candidates are left in the dark about the algorithms used and the body (or entity) behind the hiring decisions. Because data sets and software vary from company to company, there is no uniform regulation or auditing of these systems, raising considerable ethical concerns. “In terms of standards, it’s a bit like the Wild West”, laments Prem Kumar.
Although the supplier of such software is theoretically responsible for testing and implementing safeguards, this process often lacks, if not completely, independent control. Most vendors claim that AI reduces long-standing discrimination in the recruitment process, but critics remain extremely skeptical. There is no way of guaranteeing that such software does not reproduce systemic and institutional biases, particularly against women, ethnic minorities and people with disabilities. What’s more, the exact criteria used to assess candidates remain unclear.
According to Rory Mir, Associate Director at the Electronic Frontier Foundation, “there is still a lot of ground to cover in terms of regulation on this issue”. For example, in the US, the Equal Employment Opportunity Commission, which sees automated systems in employment as a “new civil rights frontier”, has only recently published guidance on how AI-enabled hiring tools should comply with federal anti-discrimination laws.
The next step is to promote informed consent by allowing jobseekers to choose AI-based assessments and data collection, while conducting regular audits. For Rory Mir, “the only real cure for AI bias is to open up the entire process”.
How can disadvantaged jobseekers stay competitive with the ‘robots’? This question has given rise to a range of online tools, such as CV scanners, skills development software and coaching platforms, which simplify the application process and save applicants time.
However, it’s important to note that these tools don’t necessarily offer a completely level playing field. “The use of AI is not inherently bad”, argues Ben Grant, head of Ramped, a recently launched personalized careers service that uses Open AI to improve candidates’ CVs and cover letters. Ramped’s mission is to make job hunting more accessible and inclusive by helping jobseekers master best practice to stay competitive in today’s market. “The days when everything was done by hand are almost over,” he explains.
Take LinkedIn, a professional social network that has a long history of using AI and has developed a set of principles over time to ensure responsible use of the technology. LinkedIn’s head of product, Hari Srinivasan, asks a single question when it comes to automated technology: “How can we help individuals seize economic opportunities?”
LinkedIn has connected its hiring ecosystem to its learning platform to help candidates gain the skills they need to land a job. LinkedIn has launched more than 100 new courses on generative AI, applied AI and responsible AI, among others.
Recruitment managers and career coaches advise against trying to fool technology. Some candidates try to use white or invisible text to list skills and qualifications they don’t have. Chad Sowash, co-host of the Chad and Cheese podcast, warns that this won’t work.
The second rule, according to Chad Sowash and other experts, is to opt for a simple CV. By simple, we mean a layout that is easy to understand. Artistic CVs should be reserved for human interaction in recruitment. To successfully pass the ATS stage, jobseekers should opt for a streamlined CV. Boxes, tables, graphics, or fancy fonts can lead to automatic rejection by the software. It is therefore recommended to use legible elements, such as Times New Roman and bullets. Some CV robots even have difficulty analyzing PDF documents.
Next, it is advisable to use a keyword simulator or an online CV analyzer. This allows you to determine how well your CV matches the job offer, in the same way as an ATS.
When we tested the software by comparing our qualifications with a randomly selected writing job advert, the keyword match rate was only 41%, well below the 70% or more range generally required.
It’s common to get a low score on the first try, sometimes because the flawed software selects keywords that aren’t really relevant. Lauren Milligan solves these automated problems by performing a process where she analyses job descriptions from various companies to determine which keywords perform best. She then uses this list of optimized keywords in place of those in the job advert. “There are ways to work with the system, but you have to understand it and use it”, she says.
These days, we delegate so many tasks to software that we forget how things used to be. Customers used travel agents to book their tickets, while others used matchmakers before the rise of dating apps.
Yet, automation always comes with human risks. In the professional world, AI poses real dangers, such as the possibility of mass redundancies and the disappearance of entire job categories.
According to Mona Sloane, a sociologist at New York University who looks at the intersection of automated technology and politics, hiring managers often don’t choose the automated hiring software their company uses and may even develop distrust of these systems after seeing their flaws. During her research, the researcher was pleasantly surprised to see how much recruiters believe the process still relies on their judgement and decision-making, and how critical they are of AI.
We asked Chad Sowash, who has more than 20 years’ experience in human resources and talent acquisition, whether personal assessment still has a place in the industry. His answer is clear: “It’s still about human interaction”.
He recommends staying authentic, telling your story honestly and answering questions in a way that matches your own experience, rather than trying to guess what the algorithm expects. He adds some tried and tested advice: practice in front of a mirror and rely on your network.
The major technology companies frequently broach the subject of AI when they publish their quarterly results. Over the past week, Google/Alphabet, Microsoft and Meta (formerly Facebook) have unveiled their financial performances, and it is notable (if predictable) how much artificial intelligence has been mentioned in the opening speeches of CEOs and other senior executives, as well as in the questions asked by Wall Street analysts.
Microsoft CEO Satya Nadella particularly highlighted artificial intelligence in his opening speech, mentioning the field 27 times. Microsoft is offering an improved version of its Bing search engine and is developing AI tools for businesses. For his part, Google CEO Sundar Pichai emphasized the importance of AI, mentioning it 35 times, in particular highlighting the power of Bard and other AI tools. Mark Zuckerberg, CEO of Meta, mentioned AI 17 times.
Mark Zuckerberg pointed out that “AI-recommended content from accounts you don’t follow is now the fastest-growing category of content in the Facebook News Feed”. He also spoke of the multiple possibilities offered by AI for connecting people, improving interactions with applications, and creating new experiences within mobile applications and the metaverse.
With regard to Bing, Nadella noted that it is “the default search experience for OpenAI’s ChatGPT, providing ChatGPT users with faster answers with links to our trusted sources. To date, Bing users have participated in more than one billion chats and created more than 750 million images with Bing Image Creator”.
Sundar Pichai also touched on the transformation of Google search thanks to AI, mentioning that user feedback has been very positive. He explained that AI was improving Google’s ability to respond to user queries, while also paving the way for new forms of search. For example, generative AI allows users to connect different dots when exploring a topic or project, taking into account multiple factors and personal preferences before making decisions such as purchases or travel bookings. This experience represents a new starting point for web exploration, allowing users to delve deeper into topics that interest them.
In all cases, a good CV should contain measurable indicators, all the personal data likely to highlight your achievements (such as the number of projects carried out, turnover generated, sustained growth, or infrastructure developed). “AI is only as good as the information you give it” ,Ben Grant reminds us. The best advice may be to accept rejections, because there will be some. And of course, don’t hesitate to attribute the rejection to the robot. But only a human being knows how to bounce back and persevere.