AI’s Next Frontier? You’re Sitting In It

Almost every AI headline this year has been about some aspect of the digital layer of work, from agents that write code, to models that take notes on conference calls. Meanwhile, the same employees who are using AI to get remarkably efficient on their laptops are walking into their real-world offices where none of that efficiency exists. 

The trouble with the AI conversation over the past couple of years is that it’s stayed almost entirely indoors, and I mean that in a very literal sense. It’s been about screens, chat windows, and documents- the stuff that happens between a person and a monitor. And all that time, the physical office has just been sitting there, treated like a fixed cost to be managed rather than a system that could actually be understood and made better.

Modern flexible workplaces, in particular, suffer from this mismatch between who’s supposed to be in and who actually shows up, which makes everything harder to manage. When employees arrive, they may not be able to find a meeting room, or they may wander the floor looking for an open desk like they’re searching for a parking spot at the mall on Christmas Eve. The floor plan itself was probably drawn up based on someone’s best guess rather than any actual evidence of how people use the space.

This inability to use the space efficiently is a bigger problem than it seems. Real estate is often a company’s second-largest expense, right behind people. And, in a recent survey conducted by The Collab Collective, six in ten workplace operations professionals said this type of workplace friction — the inability to efficiently manage office resources like desks and meeting rooms — has actually gotten worse over the past year. 

That friction has a price tag, too, and it’s significant. Employees lose somewhere between two and four hours a week to hunting for space, chasing colleagues, and working around disparate office systems. Do the math against standard compensation numbers, and a company with a thousand employees can lose as much as $9 million a year to workplace friction alone. 

The Real Estate Planning Blind Spot

AI has significant potential to address these challenges. Just as AI can forecast demand or route support tickets, it can learn which desks and rooms are being used by whom and when. It can notice where the AV equipment routinely fails and identify which floors or meeting spaces aren’t getting used. 

The old playbook for managing an office also assumed a workforce that mostly showed up on a predictable schedule. In today’s hybrid workplace, it rarely does. Occupancy data shows just how wide that swing actually is. Commercial real estate firm CBRE has found that most organizations run at full capacity on their busiest office day, but only about a third of that capacity on an average day. That gap is exactly why headcount alone is such an unreliable way to plan space. A floor built for Monday’s peak sits mostly empty by Thursday, and no amount of instinct catches that pattern the way usage data does.

The tools themselves make this worse, because most companies run booking, visitor check-in, space planning, meeting services, and analytics as separate systems that were never designed to speak to one another. A scheduling tool that has no idea how a room actually gets used can’t tell facilities anything useful when it’s time to negotiate a lease. Analytics sitting in one dashboard can’t reach over and rebook someone into a different room when their reservation falls through.

The fact is, most real estate decisions today are still largely made on instinct and on current employment, not actual usage. That approach was tenable when offices ran at consistent, predictable occupancy. It’s much harder to justify in a hybrid world where attendance swings by day of the week and by team.

Applied well, AI in the workplace does two jobs at once. For employees, it solves the logistics of working by automatically matching people to available space, surfacing what resources are free in real time and eliminating the need to move meetings on short notice. For the operations and facilities teams managing the budget, AI can analyze all those small signals — who booked what, when, how often, and which rooms sit empty — to create a clear picture of how space is actually performing. 

Even better, once a company can see how space is actually being used, managers can base their real estate decisions on real demand, just as they’d do with any other resourcing decision. With this information in hand, they can consolidate the floors nobody’s using, walk into a lease renewal with actual leverage and redesign an office layout around how their teams actually collaborate.

Where the data is unified, the payoff gets specific fast. Let’s say an organization discovered that 40% of their meetings route through a small fraction of their conference rooms. With this information, operations managers can, for instance, easily determine which rooms should take priority for AV upgrades, and which of the rest need a design rethink rather than routine maintenance.

Companies that have already put AI to work on their code, their support queues and their inboxes have already seen what happens when good data replaces guesswork. The office is simply the next place to apply that lesson. Even better, the organizations that get there first won’t just spend less on space they don’t use, they’ll also give employees back the hours they’ve been quietly losing to workplace friction all along.