Last week we brought together a group of senior workplace and real estate leaders in London for a conversation over breakfast. We were joined by Professor Prithwiraj Choudhury from the London School of Economics, one of the world’s leading researchers on digital twins and the future of work, and by Bex Moorhouse, a senior workplace strategist at WPP with more than two decades of hands-on experience managing large, complex estates.
What followed surfaced something that I think most workplace leaders already sense but have not yet fully articulated: the way we make decisions about the workplace is broken, and fixing it requires the same shift that has already transformed every other industry.
That shift is the digital twin.
What Other Industries Already Know
Raj opened with a deceptively simple observation. A digital twin takes any physical operation, layers sensors and real-time data collection on top of it, runs that data through AI and machine learning, and sends operating instructions back into the physical environment. The physical operation and its virtual replica run in parallel, continuously learning from each other.
This is not a future technology. It is running today in hospitals, factories, and energy infrastructure. Raj shared research from a hospital in New York where sensors on patients combined with AI prediction meant that half the doctors could work remotely while clinical outcomes improved, because the system was identifying which patients needed urgent attention faster and more accurately than human observation alone. At a Unilever detergent plant in Brazil, a physical digital twin now manages production entirely, with one outcome being a 6 percent reduction in gas consumption. And in Turkey, an energy company has built a digital twin headquarters in Istanbul where engineers now manage 26 power plants spread across the country from a single office. Wind engineers sit next to solar engineers for the first time. The company is transitioning from a power generation business to a technology business. Their most valuable asset is no longer the plants. It is the data network.
The pattern is the same in every case. Real-time data, AI-driven analysis, and operating instructions that flow back into the physical world. Every industry that manages physical operations at scale is moving to this model. The question is not whether it will reach the workplace. The question is how fast.
The Organizational Digital Twin
The most direct application to the workplace is what Raj calls the organizational digital twin. Most large organizations already hold everything they need: data on how individuals work and perform, how teams interact, how spaces get used, what the broader portfolio costs and delivers. The problem is that it sits in separate systems that do not talk to each other, locked away in facilities platforms, HR tools, access control systems, and finance records that were never designed to be combined.
An organizational digital twin integrates all of it into a single data infrastructure and places AI agents on top of it. Those agents can model decisions before they are made. Should we consolidate two offices? What would happen to team connectivity if we changed our attendance expectations? How would a shift in how we allocate space affect the cost base? The analogy Raj used was crash testing. Automotive companies do not crash real vehicles to test safety. They run the simulation in the digital twin. The same logic applies to organizational decisions. You run the scenario first.
This is the direction workplace strategy is heading. The organizations that get there first will make better decisions faster, with less cost and less risk. The ones that do not will keep making multi-million pound portfolio decisions based on incomplete information and instinct.
The Personal Digital Twin
Raj also shared research that I think has profound implications for how we think about leadership and organizational capacity. His experiment with the founder and CEO of Zapier, Wade Foster, produced two striking results. First, when employees received communications from either Wade or an AI agent trained on his emails, messages, and meeting transcripts, they could not reliably tell which was which. The bot passed the test. Second, employees were fine with the bot handling impersonal questions, but strongly rejected it for interpersonal ones. They do not want an AI answering questions about leadership or career. They want the human.
The implication is clear. Personal digital twins will handle the routine, the transactional, the repeated. And in doing so, they will give leaders back time for the things that actually require human judgment: the one-on-one conversation, the deep thinking, the connection. Raj pointed to Sam Liang at Otter.ai and Melanie Rosenwasser, Chief of People at Dropbox, as early examples of leaders already experimenting with this. It will not remain experimental for long.
What the Office Is Actually For
One of the sharpest moments of the morning came turning to a question that sounds basic but is still not settled for most organizations: what is the office for?
Both Raj and Bex arrived at the same answer from different directions. The office is for connection. Not the performance of presence. Not the completion of tasks that could be done at home. Connection. The deep trust-building that requires physical proximity, and the broader network-building that exposes people to colleagues outside their immediate team.
Raj’s research shows that even in a physical office, our interactions drop dramatically beyond ten meters. We circulate among the same five to seven people and mistake that for a connected culture. He also referenced a study of a company retreat where employees spent several days together and came out with more siloed communication networks than when they arrived, because people gravitated toward colleagues like themselves. The office, designed well, is the tool for interrupting those patterns. Designed poorly, it just formalizes them.
Bex brought the practitioner perspective. She described how her team at WPP shifted from tracking occupancy to tracking show-up rate, measuring how many people were expected versus how many came, and using that to right-size space and give budget relief to teams that were not coming in. She also made a point that stayed with me. In her exit interviews, every person who resigned in a recent period said some version of the same thing: I stayed longer than I probably should have for my own career because of the people. Connection is not a soft benefit. It is a retention strategy. It is a talent strategy.
Leaders Have to Go First
Bex also made the case, firmly and from direct experience, that none of this works unless leaders model the behavior they are asking for. She described a situation where occupancy was not moving despite trying every lever available, neighborhoods, booking tools, engagement events, until the team looked at what leaders were doing. They were not showing up either. Once leadership presence became a tracked and prioritized metric, everything else started to shift.
The same principle applies to AI adoption. Raj cited research from Stanford showing that senior leaders who talk about AI constantly are using it for less than one hour a week. That gap is visible to everyone in the organization and it sends a message. If you want your people to engage seriously with these tools, you have to engage seriously with them yourself.
And when it comes to the change management side of AI adoption, Bex was direct: change management is not communications. It is not an announcement email or a slide in an all-hands. It is the sustained, empathetic work of keeping people on the journey, naming the fear rather than talking around it, and being transparent about which tasks AI will change and which it will not. She shared a personal story about showing a cleaner a video of a robotic vacuum 15 years ago, thinking it would excite him. Two days later he told her he thought she was going to replace him. The technology was not the issue. The communication around it was. That lesson applies directly to every AI rollout happening in organizations right now.
What This Means for Kadence
Everything that Raj and Bex described points in the same direction. Workplace decisions need to be grounded in real data, modeled before they are made, and continuously refined as patterns change. The physical estate needs a digital twin, and the data infrastructure to power it needs to be integrated, not siloed.
This is exactly the problem Kadence was built to solve. SpaceOps gives workplace and real estate leaders the portfolio intelligence to move beyond occupancy counts, to model scenarios, to understand how space is actually being used to support the connections that matter, and to make capital decisions with confidence. The shift from measuring desks to measuring decisions is not a future ambition. It is available now.
The organizations that treat their workplace as a strategic asset, one that can be understood, modeled, and optimized with the same rigor as any other part of the business, are the ones that will get this right. The rest will keep guessing.
Book a demo with our workplace operations experts to see how Kadence is helping global organizations build that capability.
