Citadel's Talent Practice Helps Drive Space Forecasting: How People Data Builds a Better Workplace

Meet Davis and Paul

Davis Carlin is a former McKinsey colleague of mine who now leads People Analytics at Citadel, reporting to the CHRO. During a coffee catch-up, he told me something I'd never heard before in all my years of consulting to real estate teams. At Citadel, the People Analytics function in HR leads the space demand forecasting process.

Workplace, by contrast, sits under the COO and is led by Chief Workplace Officer Paul Darrah, whom I first met while he was at Google. Their charter focuses on locating, designing, building, and operating Citadel's global office portfolio. And with their HR partners providing better fidelity around the demand equation, Paul and his team can better interpret the data and translate the requirement into physical space needs.

Paul framed the difference based on his personal experience, and his network of other corporate real estate leaders:

"You run the risk, without the data, of providing the wrong recommendation. And that's expensive. This is a step change: I can make an informed recommendation to the business now, and that elevates us to strategic partner."

While CRE can sit in the HR org chart, it's often treated as a foreign body. "We can't help you, you're not HR," is what a technology firm CRE leader friend of mine recalled being told by their HR IT team — even though they report to the CHRO. To borrow from how we talk about AI rollouts, the org structure had been deployed but not adopted.

That makes Citadel's arrangement unusual. This kind of silo-busting work typically stalls because senior leaders don't want to let go of their turf.

Pressure-Testing the Demand Signal

When Davis's team first started gathering space demand inputs from business units, the signals came in across a wide range — some aggressive, some conservative, many genuinely uncertain. In-year projections tended to be grounded. Anything past the current approval horizon often stretched into a figure that is harder to validate.

Davis's team reads those signals against how hiring works. As Davis told me:

"I know how long it takes to recruit for these roles, and I know how attrition moves over the year, so I can help the business get to a number we can plan real space around."

This is where Citadel's structure differs from other organizations I've seen. A real estate team has little basis to translate a business unit's headcount forecast on its own. CRE does not own the recruiting pipeline, does not see time-to-fill data on open requisitions, and cannot translate a stated ambition into how many people can realistically be onboarded by year-end.

Davis's team can.

They own the workforce data that drives the firm's annual headcount planning, with visibility into how long it takes to recruit specific roles, how attrition shifts across the year, and where existing requisitions sit in the pipeline.

"It's not advanced calculus," as Davis put it to me. That only works when space forecasting and workforce intelligence sit together.

Stops Along the Way

For any CRE leaders reading this and thinking the Citadel model sounds impossible in your organization, I've seen two intermediate approaches in the wild.

First, Drew DePriest — then at McKesson, now at Salesforce — told me how he was using AI to make sense of the business unit slide decks and spreadsheets I described at the top of the article. This is a useful, technology-powered baby step that improves the quality of the conversation without changing who owns it.

Years earlier, a former banking client ran a Location Strategy team as a peer of CRE. Requests for sites or seats needed its approval before CRE committed resources, and that check improved how demand conversations went.

But neither approach could challenge the physics of hiring velocity.

What the New Model Unlocks

That workforce intelligence already informs how Citadel thinks about its space. Paul told me:

"The data helps us think about how we structure leases: whether it's longer term with a break option, or shorter term because we know headcount will be a little bit volatile."

Citadel's leadership also cares about what Davis described to me as "vibrancy" — the buzz and energy of a well-used floor. Paul's team has the data to measure it. Combine that with workforce intelligence on who works with whom and where the talent pipeline is heading, and you have the foundation for something most organizations cannot build: a workplace that puts the right teams near each other and keeps floors busy.

The same foundation is the prerequisite for dynamic allocation, the kind of task-based mobility I have called vibe officing.

A building that suggests where you should sit on a given day based on who you are meeting and what you are producing. A system that rebalances groups across floors as collaboration patterns shift, without waiting for the next move project to come along.

Whether Citadel goes that way matters less than the fact that it could. Most organizations cannot, because the data, analytical capabilities, and demand decisions sit in three places that do not talk to each other.

The Courage of the Move

Cross-functional collaboration is the typical prescription for this kind of problem. I have written about shared slides, unified narratives across HR, IT, and CRE, and the broader Chief of Work concept to better align data from all three.

Citadel did something different: they redesigned the work and reassigned who owns it.

The work that depends on cross-functional data now lives where the data is, and the back-and-forth between functions that normally struggle to speak each other's language gets replaced with a single team doing the work, and a better brief for the people who build.

That kind of courage is what most organizations talk about and rarely act on.

What Can You Do on Monday?

Anywhere a function evaluates requests without owning the analytical context behind them, you get a request trap or disconnected handoffs. That holds whether it is IT evaluating compute capacity, L&D forecasting training, or finance scoring project bids.

Here are three steps you can take next week:

• Audit who supplies your demand signal. If the inputs to your next major space decision come from business units in slide decks and spreadsheets, with no analytical layer evaluating them against pipeline data, your signal is unreliable by design. Find out who could supply the analytical layer.
• Move past the data-sharing baby step. Sharing HR data with CRE is the obvious first move and is worth doing. The deeper question is whether the work that depends on cross-functional data should live with the function that owns the data, or remain a cross-functional liability nobody truly owns.
• Free your builders to build. Major capital projects deserve focused leadership. If your CRE leaders spend half their time evaluating demand signals from the business, you are likely undervaluing both the demand work and the capital work. Run the math on what their time costs you.

The HR/CRE divide is usually discussed in terms of employee and workplace experience. But the real prize is bigger than cost or square footage: marrying workforce data and talent practice with what you do on workplace and real estate is how you build an environment where your best people can perform. Supply and demand runs along the same line, and costs far more to get wrong.

Give the demand question to the people who can answer it. Your workplace team gets a better brief, and your talent gets a workplace worthy of them.