When hiring, a growing number of companies now lean on artificial intelligence tools to identify top talent. They're using AI to scan resumes for keywords, rank candidates based on inferred skills based on applicants information, and even conduct initial interviews.
While it's meant to quicken the process, there's a catch: many hiring leaders don't fully trust what the bots recommend.
According to a new report from HireVue, 72 percent of talent acquisition leaders don't trust many applicants' inferred skills -- capabilities that can be assumed from information outlined on resumes. And 50 percent say they struggle to reliably confirm whether candidates actually have the capabilities their resumes suggest.
Those hiring managers want to make hiring decisions based on demonstrated skills or proof that a candidate can actually do the job, not accepting their word that they can.
The solution? It's skills-based hiring focusing on a candidate's proven abilities and practical experience rather than just their resume or credentials. According to Mike Hudy, chief science officer at HireVue, "Skills-based hiring makes complete sense, and it's where it should go," he says.
When it comes to hiring, the most important question isn't just what's listed on a resume; it's whether the candidate has the skills to succeed. With AI replacing some jobs, are those skills transferable? Can they apply them in a new role?
If your company is ready to explore skills-based hiring, keep these do's and don'ts in mind.
"AI is just a tool -- and it's only as good as the data it's trained on," Hudy says.
Right now, many hiring tools are trained to pull skills from resumes, but that foundation is flawed.
"If you're using AI to pull skills out of a resume, there are some big challenges," he explains. "There's no standardization. It's very easy just to put a prompt in and say, 'Here's a little bit about me. Give me a great resume," he says.
The 2024 Job Seeker AI Survey, compiled by the software marketplace and consulting firm Capterra polled nearly 3,000 job seekers globally and found that a majority (58 percent) are using AI tools in their current job search.
That makes it difficult to trust what your AI surfaces. Hudy points out that a high percentage of candidates embellish -- or flat-out lie -- on their resumes.
"Almost three-quarters of companies that are adopting skills-based approaches are still trying to use the resume as the foundation," he says.
The result? When companies say they don't trust AI, "I think it's not necessarily the artificial intelligence," Hudy explains. "It actually starts with the quality of the data" that the artificial intelligence is using, he said.
Instead, employers should prioritize demonstrable skills, those proven through direct assessments.
For hiring teams overwhelmed by huge application volumes, automation is necessary. But instead of relying solely on resume scanning, companies can consider embedding skill assessments at the front of the process.
These tests might include coding tests, job simulations, or multitasking exercises, all designed to evaluate candidate skills directly.
These assessments are widely accepted today and almost expected in the vast majority of jobs now on offer, Hudy explains.
Nicole Gable, Head of Recruitment Solutions for North America at global talent firm LHH, agrees that skills assessments can help with the hiring process.
"Skills assessments are quickly becoming one of the most important tools in hiring," says Gable. "Assessments give recruiters a more comprehensive, human view of what a candidate can do, which is where the rigid structure of a resume falls short."
While some candidates, especially for lower-paying, entry-level roles, may drop off when faced with a long assessment, Hudy says companies can maintain high completion rates by offering shorter, more engaging tasks.
"Generally, completion rates of an assessment are going to be 85-90 percent completion rate," Hudy adds, noting that assessments tend to perform best when they provide something of value to the candidate.
One option for providing a skills assessment is through a virtual job tryout -- an online assessment tool used by employers to determine whether a candidate can do the job by simulating real-life scenarios.
These assessments are designed to gather valuable insights about the candidate while also helping them better understand the job itself.
Throughout the tryout, candidates are given job previews and simulations that reflect what the role is really like. It might say, "Here's what someone in this job does all day -- now you try it."
The goal is to treat the candidate like a consumer in the process, giving them a realistic experience so they can decide if the job is the right fit.
Hudy said that years ago, using assessments slowed the search process down. "In today's day, it doesn't need to. In many cases, with our customers, it's helped speed the process up."
These assessments are scored quickly through automation and can even advance candidates to the next step automatically, helping to quickly narrow the candidate pool for recruiters and hiring managers to focus on.
When providing an assessment, Gable says to make sure you understand the answers to these questions: "What has a candidate learned, how easily can they learn, how have they adapted, and what are they capable of contributing next?"
In the end, "assessments offer a clearer view of the candidate as a person, such as how they think, solve problems, and adapt. When a recruiter is able to make a better match between a candidate and an open role, the entire hiring process improves," she says.
Finally, it's essential to evaluate whether your hiring process is actually working.
Hudy recommends looking at the bigger picture: Six months after hiring, how are those employees performing? Did your tools surface the right candidates? Has performance improved? Has turnover decreased?
"Ultimately, whether it's a resume, an assessment, or an interview -- the question is, is it working?" he says.
When used thoughtfully, AI can increase speed and surface hidden talent. But it needs the right foundation. That means moving away from resumes as the main source of candidate information, and instead evaluate what really matters: the skills that lead to success on the job.
At the end of the day, AI is just a tool. How well that tool works and whether it gets it right depends entirely on the quality of the data that informs it.
The opinions expressed here by Inc.com columnists are their own, not those of Inc.com.
The final deadline for the 2025 Inc. Power Partner Awards is this Friday, July 25, at 11:59 p.m. PT. Apply now.
