From sleuthing candidates not actively on the job market to stripping unconscious bias from your search, these new artificial intelligence tools will give your HR department a high-tech upgrade.
How to get more interested applicants:
Try an AI matchmaker.
- Two ex-Google engineering execs built Leap.ai, a platform that uses machine learning to analyze resumes, personal values, and job descriptions to suggest perfect-fit candidates for open roles. More than 70 percent of the people Leap puts forward make it past the first-round interview.
- Vettery's algorithms--used by companies like Peloton, Netflix, and ESPN--suggest matches from the thousands of candidates in its database. Thanks to machine learning, the more hiring managers interact with it, the smarter its AI gets about what your company wants.
Advertise the opening.
- In years past, advertising a position on multiple job boards was a manual slog. But now PandoLogic makes predictive analytics do the hard work: Algorithms use a decade's worth of historical data on millions of job ads and thousands of sites to create a targeted campaign. Then it monitors the ads in real time to tweak budget and bid rate in order to maximize views and applications.
Look to the past.
On average, companies receive 250 resumes for each opening--and many of those hopeful hires who don't land that gig could be a great fit for another role.
- In April, Google started beta testing a tool called Candidate Discovery in its Hire by Google recruiting application, aimed at small and medium-size businesses. It uses intuitive search to scan past resumes and other data to rank which past applicants might be a match for the current role.
Hunt for passive candidates.
The unemployment rate is near its lowest level in 17 years, but more than 70 percent of employees say they are either actively looking for or are open to a new gig, according to a recent Indeed survey.
Finessing that cold call.
Successfully cold-messaging a passive candidate takes skill, and most outreach falls on deaf ears. Just 5 percent of Americans have responded to a recruiter message in the past three months, according to a Textio survey.
- Textio Hire's augmented writing platform uses millions of data points on previous messages--which words and phrases worked and which didn't--to offer real-time suggestions and red flags as you type. When Zillow tested the tool, the company says the response rate for its recruiting mail climbed 16 percent.
- For those unable to fill a role, Talenya references hundreds of websites, including LinkedIn, GitHub, Stack Overflow, and Dice, as well as other public pages, to build rich candidate profiles. Then it compares those CVs with open positions--and scores the likelihood of a match.
- Wouldn't it be nice to know who is likely to jump at a job opportunity? AI platform Engage Talent uses "predictive availability signals"--like company performance, personnel changes, and news data--to calculate the odds that passive candidates might soon be dusting off their resumes.
Make hiring less of a time suck:
Employ a chatbot.
Built with natural language processing and machine learning, a chatbot can field the repetitive questions candidates have, screen for basic qualifications, and schedule interviews.
Automate the courtship.
- Mya chats up interested and passive candidates, then screens, qualifies, and sends a conversation transcript to your applicant tracking system. If you greenlight an interview (or she does), she'll find a time that works for everyone on the team, coordinate with the hopeful hire, and update everyone's calendar.
Woo via text.
- IBM's Watson technology drives TextRecruit's chatbot Ari. But humans can also step in and send personalized text messages at scale. More than a third of applicants respond to the texts within 12 minutes.
Get yourself an HR assistant.
- HiringSolved's Rai app, still in beta mode, currently communicates with recruiters, which means your hiring team can put it to work--finding candidates, refining your search, and interfacing with your email for outreach.
Let your bot roam free.
- What sets Paradox's Olivia chatbot apart is that people don't have to be in the applicant tool to engage with her. They can lob questions her way-- Why should I work at your company? What's the culture like? How's the vacation policy?--through Web, mobile, or social channels.
A better way to size up talent:
Try to clone your top performers.
- Pymetrics leads your best employees through a series of neuroscience games to gauge traits--like risk-taking, focus, and fairness--and then candidates are put through the paces. Algorithms score how closely their traits align with those of top team members. Tesla, Unilever, and LinkedIn have jumped on this gamified AI platform.
Predict the future.
- Other basic tools can scan a resume. But Uncommon, which moved out of beta in February, has a predictive element: Using a candidate's previous experience-- and redacting names, ages, and schools--the AI platform creates a merit-based profile. It ranks how well that person will meet or exceed job requirements, even if his or her resume doesn't include a given skill. Before launching Uncommon, the team trained the platform on more than 50 million resumes and six million job descriptions. Then it had companies like Amazon, Lyft, and Etsy take it for a ride. When humans reviewed the platform's picks, they were in agreement 98 percent of the time.
Put on your bias blinders.
Many hiring managers default to Ivy Leaguers, or weed out those with attributes not similar to the age or ethnic makeup of their company. Don't leave incredible talent on the table just because they don't look like you.
- Airbnb and Twitter, among others, have tested Blendoor, which captures candidate data from whatever applicant tracking system you're using, and then removes info such as names, photos, and dates. This strips out details like race, age, and gender.
- Does a female voice or a foreign accent affect your hiring decision? Not with Interviewing.io, which runs interviews using voice-masking tools for tech candidates who pass rigorous mock interviews. Twitch, Lyft, and Asana are early embracers.
Test their skills, not their talking points.
- You can set up coding interviews (even if you don't know how to write code) with Filtered's database of thousands of exercises, ranging from basic to advanced.
- GapJumpers steers job descriptions away from subjective filler like "passion" and "team player," and creates objective tests for candidates, pulling from more than 4,000 skill challenges. "This takes away the fear that hiring managers will lower the bar for quality, because they see the quality before they see the person," says CEO Kedar Iyer.
FROM THE JUNE 2018 ISSUE OF INC. MAGAZINE