Searching for people to fill jobs requires a lot of research. Experience, skill sets, salary expectations, coordinating interviews, and numerous other issues must be addressed to match worker to employer.
Artificial intelligence and machine learning are making all of those calculations easier, and the future of hiring will likely only get more streamlined due to those technologies.
“One very specific example of this is coordinating and scheduling for incoming (job) candidates,” says Dr. Andrew Chamberlain, chief economist with Glassdoor.com, a job search and salary information platform with 50 million online visitors per month. “There are companies like X.AI who now offer chatbots that handle scheduling totally.” That program, he explained, allows recruiters to shoot out meeting invites to multiple prospective job candidates and automatically fills in calendar spaces as they respond—essentially acting as an AI personal assistant. For a company that hires hundreds of new people per year, that’s a real time-saver.
Other companies, like Entelo—a San Francisco software firm whose recruitment products are used by the likes of PayPal, Target and Capital One—are using technology to help recruiters and HR personnel zero in on candidates that fit specific roles, based on skills, experience and predictive analyses that suggest whether a job candidate is more likely to be open to changing jobs, based on past behavior. “It’s supposed to make a recruiter have super powers,” says Chamberlain.
Such software can allow recruiters to overcome biases in the recruitment process. Human recruiters have shown a tendency to place perhaps too much emphasis on schools candidates attended, or even what their names are, explains Chamberlain. Software eliminates these biases. “That’s one area where these tools that do the first pass can help surface candidates based on their merit.”
Glassdoor also uses AI to help suggest the most relevant positions to job seekers, based on their search criteria, says Heather Friedland, the company’s chief product officer, which can make recruiters’ jobs easier. “We find that employers win because they are getting better-qualified candidates applying to their jobs on Glassdoor,” she explained.
Another service seeking to update the headhunting space is Talenya, a recruiting firm co-founded by Gal Almog, co-founder of recruitment tech company RealMatch (now called PandoLogic). Talenya’s technology hunts for talent by scouring the web for public information about professionals from various sites, including career destinations like Indeed.com, LinkedIn or Glassdoor; developer platforms like GitHub, online tech communities like Stack Overflow, social networks like Facebook, or even from press releases or industry articles, says Rafael Cosentino, the company’s senior vice president of development. Gathered data include job histories, job titles, location, possible salary expectations, skills, industry experience, companies worked for, and the likelihood a professional may be open to switching jobs, based on past behavior, says Cosentino. “We unify it and we create very rich profiles that we can then match with jobs.”
On top of the technology, Talenya hires and trains folks with experience in a particular industry to use the firm’s tools and act as part-time recruiters to help screen talent to fill specific roles for Talenya’s clients, which include companies like robotics firm Universal Robots, ad-tech player DoubeVerify and gift registration service, MyRegistry.com. “Ultimately, what we’re trying to do is deliver three to five amazing candidates that are interested, qualified and evaluated within two weeks,” says Cosentino.
A firm addressing the issue of bias in the job markets is Textio, which uses language to boost the effectiveness of job posts. The company’s software examines language for possible biases that may exist under the radar that could alienate certain job candidates, Chamberlain explained, identifying “overly feminine or overly masculine terms, or terms that have been associated with having trouble recruiting diverse candidates based on their past experience and their training data.” Textio’s programs also highlight phrases found to be attractive to job seekers, and suggests alternatives to problematic phrases to improve communication.
Using algorithms and software to search through and attract an army of job candidates isn’t perfect.
One weakness to using algorithms and software to search, says Chamberlain, is that savvy job hunters can construct resumes to assure theirs attract attention by stretching the truth and claiming skills or experience they know will set them apart as an AI-based culling program sorts through them. Resume padding is nothing new, but software solutions are more predictable that human recruiters in their culling behavior.
There’s also the issue of skill levels, which comes with uncertainty. “Candidates tend to list every skill they’ve ever encountered as though they’ve mastered it,” says Chamberlain. Varying degrees of aptitude can be vague and there exists no definitive way to gauge it with technology yet.
Then, of course, there are the intangibles that machines cannot pick up on, such as how job candidates perform under stress, says Chamberlain. “Do they respond with anger or do they melt down, or do they focus under pressure and perform better?” Resumes will not tell you that and, whether we like it or not, neither will AI and machine learning, now or in the near future.
“You get the false sense that we’re measuring everything about candidates and the data is telling us who is a good match,” warns Chamberlain. “You have to remember that there’s still a huge part of what we do as people that is not observable.”