What’s the future of AI in staffing?
There’s a lot of buzz in the air about AI, or artificial intelligence, in the staffing industry. But when it comes to staffing, artificial intelligence is still a work in progress—although one that’s moving quickly to help solve some of staffing’s biggest challenges.
“In many cases, AI, in our space, really just means automated integration,” says Jerome Guerard, Director of Staffing Solutions and Channel Partnerships at Monster.
“It takes the order through the client’s internal process, through the sourcing process, candidate engagement process, and then allows the last mile to be handled by a human being. The first 99 miles are covered through as many automated tools and engagement tactics as possible.”
In other words, AI is currently mostly represented by automated workflows—a combination of technologies that automates the workflow in the smartest way possible. But at present, there aren’t many situations in which a machine can take the process from start to finish. (Although there are a few. More on that later.)
And while it can create incredible efficiencies, there’s still a challenge for different technologies and platforms to integrate seamlessly. Here’s how it breaks down today and what the future might hold.
What does an automated workflow look like today?
To anyone who works in staffing, you know the drill: In most cases, a hiring manager creates a request for talent, and that order goes to the supplier or the staffing entity’s system of record.
If companies have that first tier of automation in place, orders placed in the vendor management system are auto-populated from the candidate tracking system.
If they don’t, clients must log into those platforms, find the job orders, and then manually enter the job orders into their applicant tracking system. That’s an extra step that isn’t always necessary—if you have the right tools.
“Having that first tier of automation is critical if clients want to be efficient,” Guerard says. “You now have that order in your system of record.”
From sourcing to matching
Once you’ve got your candidates in the system, AI can then pinpoint candidates within the system and use search aggregators to comb other databases for matches.
“This is where best-of-breed matching technology can instantly say, ‘You have the candidates’ or ‘You don’t have the candidates’ and if you don’t have them, here’s where you can find them,” Guerard says.
From there, a candidate engagement system might create automated engagements with the list of candidates.
With a simple AI implementation on the candidate side, user data can help refine results recruiters receive in their end. “It could be a simple question and answer decision tree,” Guerard says. “‘Like ‘Hey, we think you might be interested in this job, type 1 for yes, vote 2 for no.’ It boils down a list of 100 candidates to five candidates that would actually be a good fit for this particular position.”
The search for a one-size-fits-all solution
As it stands today, there’s no standard process that everyone in the staffing industry uses, and staffing companies vary in their adoption of different parts of these technologies, because there are so many technology partners, platforms and vendors being used.
“It’s such a complex problem that no one organization is good at doing all things,” says Ron Mitchell, CEO and founder of Virgil Holdings, a startup home to multiple AI-powered recruiting platforms. “In fact, Google just dropped their ATS product.”
The systems that will win, Mitchell says, will be those that can talk to those solutions through the process, aggregate data and, from that data, use all the information to refine their machine learning algorithms.
“No one company has all of these,” Guerard says. “But I believe we are one of the only companies that are culminating the right tech stack in which to deploy this concept in a plug-and-play manner, depending on the needs of the client.”
Where automation works best today
Although there is no end-to-end AI solution that works in all situations, there are circumstances where the technology is getting close. For specific job searches, automated solutions can use matching algorithms to find candidates, automatically engage those candidates with an email link to a chatbot interview, screen them via the chatbot and submit successful candidates for the job. This process works surprisingly well for niche software engineering roles.
“They have these really hard skills on their resume, and it matches them up very effectively,” says Chase Wilson, Vice-president of Product Innovation for Monster.
“Then the high-tech side is okay doing the chat interview, they think it’s cool, and then those people get submitted. That’s an almost fully automated find and engagement and screen process.”
But that’s one situation among thousands where AI works as a full solution. Other situations—screening for entry-level positions, for instance—don’t work as well, because the ask is somewhat more general. “We start to get a lot more irrelevant candidate resumes,” Wilson says. “Certain roles will still need better screening and more quality screening.”
Soft skills are also tough, such as for sales positions, or determining whether a candidate is a good cultural fit. “Sometimes you’ve got to meet the person and get a feel for that,” Wilson says.
How AI helps set staffing firms apart
Where AI does create opportunity, Wilson says, is in companies developing their brand—and their speed. In the end, if staffing companies are all using the same technology to select candidates, then the challenge becomes setting yourself apart from the rest.
“If I’m staffing company ABC and there’s staffing company XYZ, we may be recruiting for the same job for the same company,” Wilson says. “How do I differentiate my brand from the other brand? The staffing company that’s going to win will be the one who employs the technology effectively but doesn’t lose their branding focus in the differentiation conversation.”
And the staffing company that tops the heap will complete the process faster than everyone else. “A lot of these orders are jump balls,” Wilson says. “If I can get the candidate submitted before the next guy, then I win the day.”