How Job Applicants Try to Hack Résumé-Reading Software
Last year, Shirin Nilizadeh got a call from a friend who had been worn down looking for a job. Her friend had sent her résumé to infinite job portals, only for it to seemingly disappear into a black hole. “She was going around asking everyone, ‘What’s the trick?’” Nilizadeh remembers. Nilizadeth didn’t have job advice, but she did have an idea. A computer scientist at the University of Texas at Arlington, Nilizadeh specializes in security informatics, or the way adversaries can breach computer systems. Oh my god, she thought. We should hack in.
Most large companies use software in their hiring process. Programs called applicant tracking systems can sift through online applications and score them based on how well a candidate appears to match the open role. Some, like Oracle’s Taleo, can also rank applicants to give recruiters a short list of people to interview. The résumés at the bottom of the list can end up like those from Nilizadeh’s friend, without ever seeing the light of day.
Nilizadeh devised an experiment to see if she could trick a résumé-ranking algorithm. She collected 100 résumés from LinkedIn, GitHub, and personal websites and scraped a variety of job postings from Indeed. She then randomly enhanced some of the résumés by embedding keywords from the job posting in the text. When she ran those through a résumé-ranking program, she found their ranking improved significantly—jumping up as many as 16 spots. It didn’t matter if the résumé listed other relevant qualifications or if it appeared to match the open role.
Nilizadeh’s experiment was purely academic: She published her results last fall, with an audience of security researchers in mind. But as software pervades the hiring process, job seekers have developed their own hacks to increase their interview chances, such as adding keywords to the metadata of their résumé file or including the names of Ivy League universities in invisible text. One person, applying for an entry-level job at Google, told me they listed their Facebook page on their résumé because they believed Google’s applicant tracking systems rewarded mentions of other large tech companies. Some applicants believe such tactics help: Marco Garcia, a master’s student at École Polytechnique in France, struggled to get an interview for an internship last year until he started copying the job description of each job into his résumé in tiny white type. It was invisible to the naked eye but not to a computer. After adding the job descriptions, he told me he “definitely got more interviews.”
Sending a résumé is just one piece of the hiring process, and plenty of hiring still happens through referrals rather than cold applications. But since so many jobs are formally advertised online, recruiters rely on algorithms to wade through the flood. “You might receive anywhere from 100 to 250 résumés for a single job opening,” says Julie Schweber, an adviser at the Society of Human Resource Managers, who worked in HR for 18 years. Schweber says software can filter out as many as 75 percent of applicants who don’t meet the job criteria, and can help recruiters choose the small number of candidates to advance to the next level.
Software can also disadvantage certain candidates, says Joseph Fuller, a management professor at Harvard Business School. Last fall, the US Equal Employment Opportunity Commission launched an initiative to examine the role of artificial intelligence in hiring, citing concerns that new technologies presented a “a high-tech pathway to discrimination.” Around the same time, Fuller published a report suggesting that applicant tracking systems routinely exclude candidates with irregularities on their résumés: a gap in employment, for example, or relevant skills that didn’t quite match the recruiter’s keywords. “When companies are focused on making their process hyperefficient, they can over-dignify the technology,” he says.
“It’s more important to focus on a human looking at your résumé rather than clever tricks, like trying to stuff keywords in there.”
Nate Smith, CEO, Lever
To help workers get around these algorithmic gatekeepers, another group of companies offer to help job seekers optimize their résumés. Jobscan, one such optimizer, was founded by a disgruntled job seeker who couldn’t seem to get any interviews. For $50 a month, Jobscan offers access to software that mimics an applicant tracking system. It claims to boost candidates’ chances by showing them what recruiters are looking at, including résumé scores and keyword matching. It also suggests specific skills to add and edits out résumé clichés, like “team player” or “self-starter.” The company says more than 1 million people have used its software since it launched in 2014.
Other tools, like ResyMatch and Résunate, help job applicants see how well their skills match a job description and suggest how often they should mention specific keywords in their résumé. Austin Belcak, who created ResyMatch, says this technique works similarly to the way people tried to boost their placement in search results in the early 2000s, where they would “take a bunch of keywords and write them on their website in the same color as the background.” A visitor to the webpage wouldn’t notice, but Google would pick up on it and would boost the website’s page rank. Techniques have evolved since then, creating an entire field of search engine optimizers. Similarly, Belcak says it’s fairly simple to optimize a résumé, but some of the applicant tracking systems are getting smarter.