Artificial Intelligence Doesn't Fail. Disconnected Data Does.
AI Has Become the Center of Every Conversation. That Doesn't Mean We've Solved the Right Problem.
There is hardly a commercial real estate conference, technology event or industry publication today that doesn't promise an AI-powered future.
We're told that artificial intelligence will predict equipment failures before they happen, optimize leasing strategies, automate reporting, improve energy efficiency and transform the way properties are managed. Those ambitions are not unrealistic. Many of them will become part of everyday operations sooner than we expect.
Yet despite the excitement surrounding AI, a simple question continues to surface whenever I speak with property owners, operators and technology leaders.
If AI is advancing so quickly, why hasn't it fundamentally changed the way commercial properties operate?
My answer is usually unexpected. I don't believe artificial intelligence is the problem.
In fact, I don't think most AI initiatives fail because the technology isn't capable enough. They struggle because the operational environment surrounding that technology was never designed to support it.
Artificial intelligence doesn't fail. Disconnected data does.
Commercial Real Estate Has Never Had a Data Shortage
Every commercial property produces an extraordinary amount of information.
A maintenance request creates data. So does a contractor visit, a lease amendment, an inspection, a purchase order, a visitor complaint, an energy reading or a marketing campaign.
Individually, these events may seem insignificant. Collectively, they describe almost everything that happens inside a property.
The irony is that commercial real estate has spent decades investing in systems that capture this information. ERP platforms manage financial records. Maintenance systems track assets and work orders. Leasing platforms store contract information. Document management systems organize compliance records. Security platforms monitor incidents.
Most organizations are not suffering from a lack of information.
They are struggling with something far more subtle.
Their information exists, but it rarely exists together.
Data Alone Doesn't Create Understanding
Imagine being handed every page of a novel, only to discover that each chapter has been placed in a different room. Nothing is missing. Every sentence is intact. Yet reading the story becomes almost impossible because the relationships between the chapters have disappeared.
Operational data behaves in much the same way. Every department maintains a piece of the bigger picture. Finance understands costs. Facility management understands assets. Leasing understands tenants. Marketing understands customer engagement. Security understands incidents. Each system answers its own questions remarkably well. Very few can answer questions that span the entire organization.
That distinction is more important than it may first appear.
Artificial intelligence doesn't simply analyze information. It learns from the relationships between pieces of information. Remove those relationships and even the most sophisticated models are forced to interpret isolated events instead of connected operations.
Context Is the Difference Between Prediction and Judgment
Conversations about AI often begin with algorithms. Which model should we use? Should we deploy a large language model? Predictive analytics? Computer vision? A specialized machine learning model?
These are important technical discussions, but they are rarely the ones that determine whether an AI initiative succeeds. The more fundamental question is whether the organization has created enough operational context for intelligence to emerge.
Can a maintenance issue be evaluated alongside supplier performance? Can visitor behavior be understood in relation to marketing activities? Can energy consumption be interpreted together with occupancy, weather conditions and building operations? Can an AI assistant understand why a decision was made, rather than simply recording that it happened?
Context gives information meaning.
Without it, artificial intelligence may produce accurate predictions, but it struggles to produce reliable decisions.
Buildings Don't Operate in Silos
One of the challenges I often see in commercial real estate technology is that software tends to mirror organizational charts. Facilities are managed in one application. Finance in another. Leasing somewhere else. Security somewhere else again.
From a software perspective, this makes perfect sense. From an operational perspective, it does not.
Buildings don't experience problems department by department. A delayed maintenance activity may influence tenant satisfaction. Tenant satisfaction affects lease renewals. Lease renewals shape financial performance. Financial priorities determine operational investments. The organization may be divided into departments. The property itself is not.
Artificial intelligence needs to understand the property as a connected operational system, because that is how the real world behaves.
Integration Comes Before Intelligence
Over the past thirty years, enterprise software has become increasingly specialized. Every generation of software became better at solving a specific problem. ERP systems transformed finance. CRM platforms reshaped customer relationships. Facility management applications improved maintenance. Countless specialized tools emerged for individual operational functions.
Collectively, they delivered tremendous value. They also introduced a new challenge. Organizations became digitally capable, but operationally fragmented.
This is why I believe the next stage of digital transformation is no longer about adding more software. It is about creating a common operational foundation where existing systems can finally work together.
At EPPSO, we describe that foundation as a Commercial Real Estate Operating System (CREOS).
A CREOS is not another application competing with ERP or facility management software. It is the operational layer that allows those systems to share context, coordinate processes and contribute to a common understanding of how a property functions.
Only then does artificial intelligence gain something it has never truly had before:
A complete picture.
The Real Opportunity Is Better Decision-Making
When artificial intelligence is discussed publicly, the conversation usually gravitates toward automation. Automatically generated reports. Autonomous maintenance schedules. AI assistants responding to questions. Workflow automation.
These applications are valuable, but I don't believe they represent AI's greatest contribution to commercial real estate.
The more profound change will happen inside the decision-making process itself. Property managers will spend less time searching for information spread across disconnected systems. Executives will spend less time reconciling reports prepared by different departments. Operational teams will spend less time moving information manually between applications.
Instead, they will focus their attention where it has always created the greatest value: solving problems, improving experiences and making better decisions.
Technology should never replace operational expertise. It should strengthen it.
Measuring Artificial Intelligence by Business Outcomes
Technology teams naturally evaluate AI through technical indicators such as model accuracy, prediction quality or processing performance. Those measurements are important because they reveal how well the technology performs.
Business leaders, however, tend to evaluate success differently.
They want to know whether decisions are being made more quickly, whether departments collaborate more effectively, whether operational risks are identified earlier and whether tenants, visitors and partners experience measurable improvements.
Ultimately, the success of artificial intelligence is not defined by how sophisticated its algorithms become. It is defined by whether the organization itself becomes more capable, more responsive and better informed.
Technology creates value when it improves the quality of decisions people make every day.
Looking Ahead
Artificial intelligence will undoubtedly become one of the defining technologies of commercial real estate. Not because algorithms will continue to improve—although they certainly will—but because organizations will gradually begin connecting the operational environments those algorithms depend upon.
The companies that create the greatest value from AI are unlikely to be those with the largest technology budgets or the most ambitious pilot projects. They will be the organizations that first connect their people, processes and information into a coherent operational ecosystem.
Intelligence is never created by data alone. It emerges when information becomes connected, contextual and actionable. That is why I remain convinced that artificial intelligence is only part of the story.
The foundation comes first. Because artificial intelligence doesn't fail. Disconnected data does.
The EPPSO Perspective
Artificial intelligence should not be viewed as a feature that can simply be added to existing software. It is the natural outcome of connected operations. Organizations that first establish a common operational foundation will consistently generate more value from AI than those that focus solely on adopting newer models.