If you run talent acquisition for a health system, you already know nurses are scarce and vacancies are expensive. That's your daily reality, not a revelation. So in our conversation with Joe Gage, we went looking for the parts that don't show up in the standard pitch deck.

Joe is Chief Administrative Officer at Bon Secours Mercy Health — a top-five US Catholic health system, second-largest in Ireland, with a global service center in the Philippines and roughly 60,000 employees. He owns HR, IT, the global service center, and three schools. Here are the five things he said that a seasoned TA leader should actually pay attention to.

See why healthcare's leading talent acquisition teams recruit with Take2. Book a free demo →

1. Your funnel loses the best candidate to calendar physics, not screening rigor

Every TA leader watches conversion rates. Almost none attribute drop-off to scheduling timing — yet that's where the damage is. "You'd love to say the best hire always gets to the end of the chain, but that's not true," Joe says. "Some are falling out just because she was too busy to talk to you that night."

This is the uncomfortable reframe: your funnel isn't selecting for the best candidate, it's selecting for the most available one. Picture the nurse working "three twelves" with a family at home — the window to reach her doesn't overlap with your recruiter's workday. The candidate you lose this way never shows up as a rejection. She shows up as nothing. If your reporting can't see logistics-driven attrition, your quality-of-hire numbers are quietly distorted at the top of the funnel.

2. The volume problem is two opposite problems at once

The standard automation narrative assumes drowning in applicants. Joe's point is sharper: a single recruiter carries "60 to 70 requisitions" where some roles draw "three, 400 candidates" and others draw almost none — simultaneously. "Speed in places where there are a few is critical, and then being able to sort quickly in places where there are a lot is important."

That means a recruiter has to run scarcity tactics and flood tactics on the same desk, the same day. Most tooling optimizes one or the other. The real design requirement is a system that compresses time-to-touch when candidates are rare and compresses time-to-sort when they're abundant — without the recruiter context-switching between two mental models all day.

3. Human capacity can't flex to demand spikes. At all.

Leaders tend to frame AI as a cost-savings story. Joe frames it as a capacity elasticity story, which is a different and more strategic argument. He recalls a vacancy spike: "It literally took me months to hire the recruiting coordinators I needed. And then they started working on the backlog... it was a several-million-dollar exercise to beat down a spike in vacancy."

The insight isn't "AI is cheaper than a coordinator." It's that human capacity has a hiring-and-training lag measured in months, so it structurally cannot respond to a demand spike when the spike happens. By the time you've staffed up, the spike is over and you've over-hired. Agentic systems absorb volatility on day one — "the tech does not worry about working overtime." For a leader managing seasonal and crisis-driven swings, that's the part worth modeling.

4. First-year turnover is a morale contagion, not just a cost line

You can quantify the $40K–$60K replacement cost — and Joe notes "I have yet to sit in a room with a CFO who has questioned that number." But the cost he wants leaders to weigh is the one that doesn't fit in a spreadsheet. A bad hire becomes "almost a year-long mistake," and the churn that follows is "demoralizing for those that work in that unit because they're watching one person after another come through and not join as a successful member of the team."

Turnover compounds. Each failed hire degrades the stability of the unit that has to onboard the next one — which makes the next hire likelier to fail. It's a retention problem disguised as a recruiting problem.

5. This is the rare case where quality and cost don't trade off

Use this one in your next budget conversation. Joe makes the argument that a stable workforce is simultaneously the best clinical outcome and the cheapest one: "What's for the patient and what's good for the hospital and what's good for the health insurer and what's good for the employer are all the same thing. A stable workforce is the best thing."

When you're defending headcount or tooling spend, the usual framing pits cost against quality. Here they point the same direction — which is a stronger position than most TA leaders give themselves credit for.

Where the line actually sits

Joe is precise about what to automate and what to protect. Data collection, scheduling, credential verification — "those are perfect for automation." But the moment trust gets built stays human: "What's it going to be like to work for you as a boss? Are we going to connect? That's where trust is built eyeball to eyeball." Automation's real job is to clear the toil "so that you can maximize the points of human connection."

His parting line for the skeptics is the strategic stakes in one sentence: "If you're not using this tech to be fastest and first to the candidate, your competitor will."

See how Take2's AI agents can transform your recruiting operation. Book a 30-minute call. We'll walk through projected impact based on your role types, requisition volume, and recruiter capacity.