John Vlastelica has spent his career on the outside of talent acquisition teams looking in, which turns out to be a pretty good vantage point. As CEO and founder of management consulting firm Recruiting Toolbox, he's worked with organizations across nearly every industry, watching how they build their hiring processes and, increasingly, how they're rethinking those processes around AI.
We recently sat down with John to chat about what he’s learned from his firsthand knowledge of the industry’s evolution, and how companies can get the most out of the tools they use.
As automation has become more common in the hiring process, how do companies differentiate themselves from one another?
John Vlastelica: As a consultant, I've worked with hundreds of organizations, and one thing I keep seeing over and over again is convergence. Companies are buying very similar products within the same categories, and there are usually one, two, or three clear leaders that most TA leaders land on. When I get a room full of heads of TA together and do a show of hands, a lot of them are using the same handful of tools.
When I get a room full of heads of TA together and do a show of hands, a lot of them are using the same handful of tools.
So differentiation isn't really about which platform you pick anymore. Differentiation ends up coming down to the human piece. One company might decide not to automate a certain step because they want to insert a high-touch moment there instead. Another might bring the hiring manager in at a specific point so there's an escalation option built into the process. Those choices, more than the technology itself, are what actually separate one company's hiring experience from another's.
When you implement new software, what tests do you use to measure whether it’s actually working?
JV: You can A/B test almost anything now, including your hiring process. Let’s say we want to track two different groups of candidates. We’ll run one group through a fully automated, conversational AI experience, and we’ll run the other through an experience where the hiring manager stays hands on and meets every candidate. Then we’ll measure the difference. To that point, speed is the easiest thing to track. Quality, on the other hand, is harder to measure, but retention or time-to-onboard actually serve as decent proxies, even if they take longer to show up. Some of the friction I see is simpler than that, though. Candidates get stuck just trying to search for open roles. They see 17 openings with the same job title in the same location and have no idea which one to apply to. Fixing that kind of confusion is its own kind of test — and usually requires some digging into your own systems — but it's often worth as much as the fancier automation experiments.
Beyond measuring it, what's the hardest part of actually getting people to use it?
JV: The hardest part is never the search. It’s never the RFP. It's change management: getting people to actually change their behavior. Granted, good technology makes change management easier. When an AI hiring tool eliminates the ‘crap work’ that recruiters have had to deal with, it frees them up to focus on higher value work. And that shift changes how TA interacts with the rest of the business.
Differentiation isn't really about which platform you pick anymore. Differentiation ends up coming down to the human piece.
Before, hiring managers would have to hear, “Sorry, we're too busy clicking buttons.” When a team automates the tedious hiring work, they instead say, “We can react to this need you have.” That's the real ROI: you’re able to say yes to new kinds of requests.
Do you think TA leaders are going to start building more of their own tech?
JV: I think about this a lot. Is vibe coding going to get decentralized down to the point where directors in TA are building some of their own stuff? I don’t know. What I do think it speeds up is getting aligned on what a team actually needs. That alignment conversation moves a lot faster when someone can show you an idea instead of just describing it. Handing that idea off to a vendor to actually build and secure at an enterprise level is a great model. But the idea of fully vibe coded solutions running at enterprise scale, with all the security that requires, makes me pretty nervous right now.
What has you most excited about where this all is headed?
JV: Everything is just moving so quickly — voice is a great example of this. Take CarPlay. You can be driving down the road, realize you’re lost, ask your car for directions to where you’re headed, and it works. That kind of natural voice interaction felt clunky even a year or two ago.
What I love most about all this progress, though, are the stories on the other side of it. Someone's been searching for a job for three months, an agent pops up and asks what they're actually looking for, matches them to a role, and they get hired. Those are the stories worth telling, especially in this economy. Not the resume robot doom and gloom, but AI actually helping someone get unstuck.
