Artificial Intelligence’s New Role in Recruitment
Nowadays, many companies use artificial intelligence to sift through resumes and create a shortlist of stellar contenders. But did you know many companies use recruitment AI to find promising candidates and advertise jobs directly to them? AI can also be used to maintain consistent text or email communication with job candidates throughout the recruiting process. Some software is even able to predict soft skills and job performance, or how likely a candidate is to accept a specific job position.
The recruiting process has changed drastically in the last couple of decades. Thankfully, this blog is here to help you keep up with the latest technology transformation for talent recruitment.
In this article:
- Why hiring managers are turning to AI
- Explaining application tracking software (ATS)
- How LinkedIn and Indeed use AI
- Conversational AI and natural language processing (NLP)
- Reducing bias in recruiting
- Catering your resume to ATS
- Soft skills measurement software
Why Are Hiring Managers Turning to AI?
Online job applications aren’t new. In the last decade, platforms like LinkedIn, Indeed and Monster have skyrocketed in popularity. In fact, many employers are no longer accepting mail-in or in-person job applications.
Virtual recruitment practices have continued to progress, with online application software and career networking platforms becoming increasingly complex behind the scenes. This growing complexity is enabled by artificial intelligence, which is working behind the scenes to make the recruiting process easier for recruiters and job candidates alike.
The increasing ease in application tools means more applications are making their way to employers. However, it has also made it harder for candidates to stand out from the crowd. It also means that manually sifting through resumes is almost impossible for hiring managers. As a result, employers use application tracking software (ATS) to quickly screen applicants to find the most well-suited for the job, ensuring that they don’t miss the most promising candidates.
Being aware of AI driving the recruitment process impacts our ability to get the job we want by helping job applicants and recruiters to make better decisions when applying for jobs, recruiting candidates, and hiring new employees.
As technology becomes more integrated into every single industry and jobs become more specialized, a lot of companies are struggling to find the right talent. The ease of online applications means that employers are getting bombarded with unqualified applicants. Artificial intelligence plays a big role in connecting recruiters to the right candidates, whether it be through application tracking software (ATS), recruitment platforms like LinkedIn and Indeed, or online skills assessments.
Instead of taking advantage of online job postings to send your resume to as many jobs as possible, try tailoring your resume to recruitment AI to accurately showcase your value to a company. While it might feel like a drag, putting in the hard work now is the best strategy in the long term.
What is ATS? How does it work?
ATS is a type of software that manages a company’s recruiting and hiring processes. Employers can use it to organize candidates and their information, as well as filter job prospects according to different criteria. ATS can use info from previous successful applicants to identify suitable candidates and even become optimized to a company’s hiring preferences over time (Source: Vox).
Overall, ATS helps companies reduce time-to-fill (the time it takes to fill a vacant job position) and save recruiting costs. That’s why over 98% of Fortune 500 companies use applicant tracking systems (Source: Jobscan).
Think of ATS as automating the resume skimming process that a human recruiter would perform. Recruiters are provided with a helpful AI-generated shortlist of candidates, cutting down on paperwork and allowing more time for face-to-face interactions with candidates.
How Recruitment Platforms Use AI
Nowadays, most job recruitment platforms use some form of artificial intelligence with the goal of creating successful, mutually beneficial hires. AI helps them connect job seekers to employers and vice versa.
How LinkedIn uses AI
LinkedIn uses AI to give employment recommendations, suggest connections, and show users relevant posts. The mantra behind their feed is “People you know, talking about the things you care about” (Source: Analytics Insights). So how does the platform determine what people and topics are relevant to you?
LinkedIn uses your activity, combined with information inputted by other users, as inputs for their algorithm. All these inputs are compiled to build LinkedIn’s “knowledge graph”, a giant database of every member, job, title, skill, company, location, and educational institution (Source: LinkedIn Engineering). The knowledge graph is updated in real time using machine learning to grow and scale as information is continually added to LinkedIn by members (Source: LinkedIn Engineering). As a result, the graph’s interpretation of the relationships between people, jobs, and other elements doesn’t stay static but is constantly changing to reflect reality as accurately as possible.
LinkedIn uses the data found in the knowledge graph to group people, jobs, and other elements into “entities”, which are groupings of people, job titles, and other elements with large overlap (Source: LinkedIn Engineering). This means that the posts and jobs that appear on your feed aren’t just based on your activity, but also on the activity of users with profiles similar to yours. On the recruiting end, grouping job candidates and their titles and experiences into entities allows them to find new talent pools that use keywords they weren’t necessarily searching for (Source: LinkedIn Engineering).
One issue with user-generated data is that different people and companies use different terms for the same things. One job description could use the title “software engineer”, whereas another uses the title “computer programmer”, even though these positions are similar and would be relevant to the same pool of job applicants. LinkedIn uses machine learning models to infer your skills based on your experience and show you the most relevant job opportunities, despite possibly mismatched keywords (Source: LinkedIn Engineering).
The efficiency of their algorithm is measured through relevance metrics, such as the number of members clicking on job listings. Additionally, LinkedIn uses A/B testing to test their algorithm and continually improve it (Source: LinkedIn Engineering).
How Indeed uses AI
Thousands of jobs are posted on Indeed each day (Source: Indeed), and 250 million job seekers visit their website every month (Source: Financial Post). When job seekers search for a job on Indeed, the platform uses AI and machine learning to filter results based on the data they’ve collected and help job seekers find the opportunities most relevant to them (Source: Financial Post). Think Google or Youtube’s search results algorithm, but for jobs.
Indeed’s platform uses natural language processing (NLP) to “read” job descriptions and resumes, pinpoint important information, and create better matches for recruiters and job seekers alike (Source: Indeed for Employers). Additionally, through AI and machine learning, Indeed makes informed assumptions about jobs being posted. For example, if a job description is unclear or contains gaps, Indeed’s algorithms can step in to estimate salaries or classify unusual or eccentric job titles (Source: Indeed for Employers). That being said, optimizing your professional networking profiles and keeping them up-to-date might be the key to accessing more tailored job opportunities online.
Conversational AI Is Changing the Game
Nowadays, 64% of people would rather message a business than call them (Source: Facebook IQ). Conversational AI helps employers stay in contact with prospective employees and keep them interested in a job. It also allows them to display their company culture and create a good impression with customers.
Conversational AI combines natural language processing (NLP) with software such as chatbots and voice recognition (Source: IBM). NLP deciphers the meaning of user-inputted data (text or speech), formulates a response, and then uses machine learning to verify results and improve the accuracy of responses over time. As more data is added, artificial intelligence becomes more adept at recognizing meaning and generating correct responses. This allows AI to collect info and ask screening questions in a conversational manner, as well as interpret user answers, ensuring that recruiters demonstrate responsiveness to candidates and aren’t missing out on top talent (Source: XOR).
Reducing Unconscious Bias in the Recruiting Process
One of the main advantages of AI in recruiting is that it helps reduce unconscious bias. Recruitment AI relies exclusively on data points to determine the best candidates for a role, as opposed to a person who may be unconsciously biased. However, since AI is completely impartial, it can sometimes detect causal relationships between unrelated elements and unknowingly deliver discriminatory results when left unregulated.
In 2015, Amazon scrapped their attempt at creating an AI capable of pinpointing the top candidates for any role. The AI pulled out top resumes based on data from previous successful resumes received by Amazon. Since most of these resumes belonged to men due to the tech industry’s gender gap, the algorithm was based on an imbalanced data set and the AI trained itself to penalize resumes featuring the term “women’s”, as in “women’s college” or “women’s team” (Source: Reuters).
This type of preferential treatment is why it’s necessary for recruiting AI to go through regular audits. Recruitment AI is often trained using data from past successful candidates, which can introduce bias into the algorithm when AI picks up the wrong similarities as causal (such as focusing on gender, as opposed to knowledge in coding). Therefore, the ideal solution to automating recruiting isn’t to eliminate the human component, since it is needed to analyze AI and algorithms and assess the quality of results. Instead, we need to aim for the right marriage between automated recruiting with AI and professional oversight.
The Government of Canada recommends reducing bias in the recruiting process through approaches such as “blind” evaluations, skills and knowledge testing, standardized interviews and “de-gendering” job ads (Source: Government of Canada). Artificial intelligence can help recruiters implement these strategies by automatically removing identifying information from job applications and conducting virtual skills assessments.
CENGN alumni Fintros is a talent discovery software for finance professionals that focuses on removing hiring bias in the finance industry. They invented the Inclusive Resume™, which uses AI to automatically anonymize your profile and send it to top finance companies without identifying information. The Fintros platform is complete with compliance monitoring to ensure that companies are hiring based on merit.
Learn how Fintros helps professionals find finance career opportunities.
How to Make Your Resume ATS-Friendly
While ATS and platforms like Fintros help mitigate bias in the hiring process, they can’t do all the work. Making sure that your resume can be scanned and understood by applicant tracking software is an essential step in the online job application process.
Here are 5 tips for creating a more ATS-friendly resume:
1. Use keywords.
It’s important to carefully read through the job description and pick out the core competencies required for the role. These keywords can typically be separated into hard skills and soft skills.
Keywords are specific words or phrases that describe the skills the employer is looking for. When using an ATS, companies often input specific keywords for recruitment AI to scan for in resumes. If your resume does not contain these keywords, it will be overlooked by ATS and will never make its way into the hands of a recruiter.
Look for the hard skills that come up more than once in a posting and are mentioned near the top of the requirements and job duties. They are the abilities, skills, expertise, and values that the recruiter is looking for in a candidate (Source: Novorésumé). Hard skills are more concrete and easily quantifiable. They include types of software, methodologies, or languages that a candidate needs to know.
Soft skills are skills like critical thinking, problem-solving, and communication. These are also often listed in job descriptions, and the ones which fit your experience and personality can be added to your candidate summary.
If you’re asked to list your skills, do your research about what specific keywords your industry is looking for. It’s possible that companies you’ve worked for in the past use terminology that isn’t used by a large portion of other companies, and that you’re missing out on job opportunities by missing those keywords.
2. Spell out acronyms
To make sure that your skills and certifications are picked up by an ATS, it’s important to spell out acronyms like B.A., MBA, and CPA, then puts acronyms in parentheses. This way, your skills or education level will be picked up by ATS whether the software has been programmed to recognize the long-form or shortened version of the word.
3. Make sure your resume format isn’t overly complicated
While previously, fun graphics or an interesting resume layout may have differentiated you from other candidates, in the age of AI, they can put you at a disadvantage. Clean and simple resume formats are preferred. ATS is usually unable to read images, tables, text boxes, and text in columns, so it’s best to avoid them altogether (Source: Career Center). Moreover, don’t put text in the header or footer section. If you want to add your contact information as a header, consider decreasing the top margin of your document (Source: TopResume).
It’s also important that the actual information in your resume is presented in a standard format. The safest route is to list your job experience in reverse-chronological order, with your most recent experience at the top. If you list experience in a different order, ATS will likely be unable to properly recognize the information in your resume.
4. Use a standard font
Similarly, it’s preferable to use a standard font for your resume. This prevents a situation in which ATS cannot read your resume because it cannot read your font. Additionally, this makes it easier for a human recruiter to read or skim your resume, without distracting from the actual content of your resume.
Standard fonts that are recommended for your resume include Calibri, Georgia, Times New Roman and Arial.
5. Remember that people will be involved in the process later
When putting your resume together, it’s important to not exclusively focus on ATS and to balance catering to ATS with catering to recruiters.
There’s no point in using cheating tactics to boost your resume for recruitment AI when a traditional recruiter will be able to see through tricks. For example, keyword stuffing is the practice of loading your resume with irrelevant keywords to manipulate its AI ranking, even if they don’t make sense within the context or accurately reflect your experience. Even if these types of cheating strategies could get your resume past ATS, the human recruiter who will look at your resume afterwards will immediately dismiss it.
Make sure the keywords you are lifting from the job description apply to you so that you attract human recruiters who are looking for someone with your skillset.
Looking to give concrete proof of your cloud computing skills on your resume?
Put Your Resume to the Test
There are a variety of websites that scan your resume using artificial intelligence and provide actionable items, such as Jobscan, Resume Worded and ZipJob. These websites have paid versions that unlock more detailed advice, but their free versions still provide helpful feedback on the effectiveness of your resume format, writing style, keywords, and action verbs. Try using a combination of resume scanners to gain deeper insights into the effectiveness of your resume when put through ATS.
Measuring Soft Skills Using AI
Worried that the focus on resumes and keywords isn’t giving employers an accurate overview of what you bring to the table? Some recruiting AI is tasked with de-emphasizing resumes and giving recruiters a more holistic view of job candidates.
The best candidate on paper may not actually be the best fit for the job. Some people may have followed an untraditional career path but be eager learners and a perfect fit for the company culture.
Skills assessments are one way of delivering a more well-rounded view of candidates and preventing bias in the recruiting process.
CENGN alumni Nugget.ai leverages the power of AI to help recruiters match people with jobs based on hard and soft skills. Job prospects are asked to complete skills assessments, after which nugget’s platform uses language processing and pattern recognition to identify discrete behaviours that are translated into soft skills. Their software is even able to use assessments to give an overview of a team’s soft skills and identify which job candidates would help balance out their skills.
Check out how Nugget.ai uses AI to hire, train, and develop top talent.
Recruitment AI and Job Seekers Working Together
In such a competitive job market, companies and job candidates alike need to stay up to date on the latest HR technology to create the best job matches. AI has ingrained itself in the recruiting process, optimizing the job search process and giving employers a fuller view of promising candidates, beyond their concrete skills.
ATS helps recruiters take on the overwhelming hoard of applicants. This is why, for job seekers, the best way to get a job isn’t to blast out your resume but to target specific positions relevant to your experience and cater your online work profiles and resume to ATS and job search platforms.
Using AI technology correctly could be the solution to your recruiting needs, or even landing that big next job for your career.