There is a version of commercial real estate management that operates entirely in the past. Owners receive reports, review what happened last month, adjust next year’s budget by a small percentage, and repeat. It works, until it stops working.
Bill Douglas, CEO of OpticWise, has spent the last decade making the case that the commercial real estate industry is running almost entirely on reactive data. And in a market where rent growth is slowing and operating costs keep climbing, that approach is becoming harder to sustain.
“Do you want to look at the results of your reports, or do you want to look at the root cause of what is driving your indicators?” Douglas asks. “Do you want predictive analytics or reactive operations?”
For most portfolios today, the honest answer is reactive, whether they intended it that way or not.
The Problem With Waiting for Reports
When a property’s numbers decline, the typical sequence is: the property manager notices, generates a report, sends it to the home office, and six weeks later an asset manager asks what happened. By then the damage is done and the response is expensive.
This is not a failure of process. It is a failure of access. The data that would have predicted the problem was available all along. Utility demand curves, space utilization patterns, equipment performance trends, these signals exist inside the building systems every single day. But because that data never leaves the vendor platforms it lives in, the people responsible for portfolio performance never see it.
“Operating technology is often the root cause of poor financial outcomes,” Douglas said. “Do you have the ability to look at that in depth? Do you have the ability to apply AI to that data set?”
For the vast majority of commercial real estate portfolios, the answer is no.
What Predictive Actually Looks Like
Predictive analytics in this context does not mean sophisticated AI dashboards and real-time alerts on day one. It starts much simpler than that.
It starts with collecting operating data for six to nine months across your key systems. Lighting, HVAC, access control, parking, leak detection. Once you have that history, you can begin running machine learning across it to find correlations a human analyst would never catch.
Douglas gave a concrete example from utility management. A large building’s air conditioning system typically starts up in response to a temperature threshold. Simple thermostat logic. But if you have data on occupancy patterns, event schedules, and peak utility rate windows, you can instruct the system to delay startup by seven minutes and avoid the demand charge entirely. That one change, across a large commercial property, can save thousands of dollars a month.
“It is not difficult to do,” Douglas said. “It is very tedious, and it takes a plan.”
Another example: a building OpticWise audited had a lighting control system installed six years earlier that had never been turned on. The system was paid for, the monthly software fee was being paid, and nobody had ever activated it. Once activated, the client saved $70,000 in utility costs that year alone.
That is reactive management at its most costly. The system existed. The savings were sitting there. Nobody knew.
The Data Exists. Owners Just Do Not Have It.
One of the most common misconceptions Douglas encounters is the belief that building data does not exist or would be hard to collect. In reality, the data is already being generated. Every system in the building is producing it. The problem is that it lives in vendor platforms, not in the hands of the property owner.
“They own it, but they have not collected it,” Douglas said. “They have let their vendors have it and manage it, and they have never gotten it. It is like owning a car on the other side of the country and saying you could drive to the store. You can’t.”
Getting that data into a format where it can be used for predictive analytics requires a deliberate strategy: identify what systems you have, understand what data they are generating, build a data lake that pulls it together, and then let machine learning surface the patterns.
The Peak Property Performance book by Douglas and Drew Hall, published by Fast Company Press, lays out a five-step framework for exactly this process. It is designed for owners and operators, not technicians.
Why the Shift to Predictive Matters Now
The market context makes this more urgent than it used to be. Rent growth has slowed significantly. OpticWise data suggests the market may only tolerate increases of one to one and a half percent this year. That means the path to stronger NOI runs through expense reduction, not revenue increases.
Utilities, insurance, and occupancy are the three biggest levers. All three can be moved with better operational data. None of them can be effectively managed from a summary report.
Owners who build optimal data & digital infrastructures now will have a compounding advantage. Machine learning gets more accurate over time. The correlations it finds in year two are richer than year one. The savings compound. The property becomes more autonomous, more efficient, and more resilient to market pressure.
“Once you can apply machine learning to your data, you can drive expenses down and income up,” Douglas said. “It is unbelievable what our clients are doing with their own data. Because they have it.”
About OpticWise: OpticWise is a digital infrastructure and data strategy firm focused on commercial real estate. The company helps owners and operators take ownership of their building data, reduce operating expenses, and build the foundation needed for AI-ready properties.
This article is based on information provided by the expert source cited above. It is intended for general informational purposes only and does not constitute legal, financial, or real estate advice. Readers should conduct their own research and consult qualified professionals before making any real estate or financial decisions.