For years, the business came to Talent and HR for answers. What are our run rates? What does comp look like? Should we open a new geo? Does your team have the capacity to hit the plan? AI changed the questions faster than anyone could update the answers. Now the AI bill has reached Finance, and every role in next year's plan costs something different than it did last year.
Beverly Carleton, VP of People, Strategy, and Operations at TeamOhana, has spent 30 years in talent. Will Barnett, Director of Talent at Rad AI, helped scale early-career recruiting at Facebook from a company of about 1,000 to 80,000, and now runs talent at a 200-person AI company. Recorded live at Density SF, they make the case that Talent leaders earn their seat this year by asking better questions, and they share the ones to bring to Finance, IT, and hiring managers before headcount locks.
Attendees will learn:
- Why the 2027 conversation moves from "what's your stack" to "what did it cost, and what did it return"
- How to factor AI spend into the total cost of every hire before the req opens
- How to push back on a headcount request by redesigning the role, the level, or the mix
- The three questions to ask Finance and IT before you hire against next year's plan
Key Highlights
Talent and HR leaders built their credibility on having answers. This year, the most useful thing they can bring to the business is a question. Beverly Carleton, VP of People, Strategy, and Operations at TeamOhana, and Will Barnett, Director of Talent at Rad AI, sat down at Density SF to talk through what changes when the AI bill reaches Finance and every new role carries a cost no one has priced. Will draws on 12 years scaling recruiting at Facebook and two years running talent at Rad AI, where AI is two-fifths of the company name. Beverly brings 30 years in talent and a clear prediction for 2027 planning. Together they cover how to price AI spend into a hire, when to trade several mid-level reqs for one senior one, how to get workflow data out of engineering before headcount locks, and the three questions to ask before you hire against next year's plan.
What you'll learn
- Why Talent and HR leaders add more value this year by asking questions than by having answers
- How to calculate the total cost of a hire when AI tools add $250 to $2,500 a month per person
- How a single source of truth for headcount helped Facebook hire 30% more people than expected in a quarter
- When to trade several mid-level reqs for one principal-level hire
- How to get workflow data from engineering before planning locks, and why Talent belongs on the AI committee
- The three questions to ask before you hire against next year's plan
Takeaway 1: The business needs your questions more than your answers
Talent and HR leaders were trained to show up prepared, with run rates, comp bands, and capacity numbers ready. AI has made many of those answers unknowable for now. Beverly's point is that this is the opening. When a hiring manager or CFO brings Talent a request, the strongest response is often a question back: what will this cost, what will AI handle, and how will we know it worked? Asking those questions early protects the people you hire, the people you have, and your own team. Waiting for perfect answers leaves the workforce plan to whoever sees the invoice first.
Takeaway 2: Every req now carries an AI cost, so price it before you open it
A new engineer might spend $250 a month on AI tools, or $2,500. Multiply that across ten hires and it changes the budget. Most hiring managers haven't factored it in, and many don't know it yet. Talent sits close enough to Finance to ask. If the total cost of an employee is wrong going into workforce planning, the company finds out a quarter later, over budget, and the most recent hires pay for it. Will reports to his CFO, and he treats that link between people, performance, and finance as the job, not a side benefit.
Takeaway 3: Level for impact, not years of experience
Leveling guides mostly measure tenure. That may stop tracking impact when an early-career hire fluent in AI can orchestrate dozens of agents, or when a VP's scope looks different because agents handle part of it. Will's approach is to get deep into what the team needs before writing the req. Sit in on performance reviews. Look for patterns. Then make the case: this team may not need two IC5s, it may need one principal who can change how the team builds. Shaping the org this way up front avoids the people debt that comes from hiring to whatever the manager asked for that morning.
Takeaway 4: If no one owns the AI spend, don't hire against it
Beverly's three questions close the session. How will AI spend impact your headcount? If Finance or IT can't answer, you aren't ready to hire. Who determines AI spend? The CIO, CHRO, and CEO may each think they own it, and if no one knows where the money comes from, it shouldn't be spent. Who manages the spend? Without someone watching it by team, person, and output, AI spend grows unchecked, and the teams you just hired absorb the impact. Will's addition: look forward, not back. Last quarter's report won't help you plan next quarter's hires.
"How will AI spend impact your headcount? If they don't know the answer, you're not ready to hire yet. You can say that. You have permission. If you don't know what this is going to cost, we probably should not be hiring it." — Beverly Carleton, VP of People, Strategy, and Operations at TeamOhana
