A property company will tell you, without hesitation, what business it is in. Buildings. Units. Square footage. Leases. The physical asset is the thing you can stand in front of, insure, depreciate, refinance, and sell. It sits on the balance sheet, in a line item, valued to the dollar. That is the business. That has always been the business.

And yet almost every asset-heavy industry that arrived at this question before property was eventually forced to answer a more uncomfortable version of it. Not "what do we own?" but "where does our advantage actually come from now?" When those industries looked closely, the answer had quietly moved. It had migrated off the physical asset and onto the data about it - how the asset behaves, what it costs, who uses it, and what happens next. The machine, the aircraft, the store, the account: each became a somewhat interchangeable input. The accumulated knowledge about it became the moat.

Property is now walking toward that same line, and the purpose of this piece is to explain why the crossing is not a strategy option a company gets to accept or decline. It is closer to gravity.

Every property company eventually becomes a data company - not because it chooses to, but because the economics of the business keep rewarding data capability until data capability becomes the main event. The only real decision a property leader gets to make is which kind of data company to become: a competent one that treats its data as an appreciating asset, or an incompetent one that generates exactly the same data and throws it away.

The Pattern Is Older Than PropTech

The clearest way to see where property is heading is to watch industries that already made the trip.

John Deere is the textbook case, and it is instructive precisely because nothing sounds less like a software company than a 190-year-old tractor manufacturer. Deere's machines still move dirt, and equipment sales still anchor the business. But its competitive advantage increasingly comes from what surrounds the machine - precision-agriculture software and the operational data it captures: soil readings, yield maps, application rates, and machine telemetry streamed continuously into a platform that turns farming from intuition into a data-driven discipline. Strategy scholar Michael Porter saw the shape of this a decade ago. In his 2014 Harvard Business Review analysis of smart, connected products and how they transform competition, he argued that connected assets force companies to confront a founding question many had never had to ask: "What business am I in?" Deere's honest answer is no longer only "tractors."

It is not an isolated story. Banks discovered they were risk-and-data businesses that happen to hold deposits. Airlines learned that their loyalty and pricing data could rival the flying as a profit engine. And Amazon's advantage is widely attributed less to its physical footprint than to the logistics, data, and software capabilities behind it. In each case the physical assets stayed important - but they stopped being where the advantage lived.

The macro numbers make the migration hard to dismiss as anecdote. The long-running Ocean Tomo Intangible Asset Market Value Study tracks what actually drives the market value of the S&P 500. In 1975, tangible assets - property, plant, equipment, inventory - accounted for 83% of that value. By the end of 2025, the relationship had completely inverted: intangibles, the category that includes data, software, and intellectual property, made up roughly 92%, with tangible assets down to just 8%. Ocean Tomo calls this an "economic inversion" - value migrating, as they put it, from what can be touched to what can be thought.

The mechanic underneath every one of these transitions is the same, and it is worth naming plainly because it is what is now arriving in property: the physical asset commoditizes, while the data about the physical asset compounds. Two owners can hold identical buildings. Only one of them holds fifteen years of clean, structured, continuous knowledge about how those buildings actually perform - and that knowledge is the part that cannot be bought on the open market.

Why Property Is Next, and Why "Eventually" Now Reads as "Now"

Property has a quiet advantage that most industries envied on their way through this transition: it is, structurally, one of the richest data-generating machines in the economy. Every unit has a maintenance life. Every lease carries a payment and behavior record. Every building accumulates years of cost, energy, occupancy, and turnover patterns. This data is produced whether or not anyone captures it. The property runs; the record either gets kept or evaporates.

What has changed is that several forces are now converging to convert that latent data into the actual basis of competition. Four of them matter most.

• The first is artificial intelligence, which does something counterintuitive to the value of data. As increasingly capable foundation models become widely available to every operator, the model stops being the differentiator and the proprietary data becomes it - a point we develop in depth in your AI is only as smart as your data history. Models are becoming broadly accessible; a proprietary operational history is not. In a world of increasingly commoditized intelligence, the data is the scarce input.

• The second is institutional capital. As professional investors underwrite real estate operations and not just the bricks, they increasingly value operational transparency, reporting quality, and demonstrable operating performance - all of which depend on reliable operational data. The operating data is becoming part of what gets diligenced.

• The third is margin. Data-driven pricing, renewal, and maintenance decisions move net operating income, and in real estate even a one-percent NOI improvement can meaningfully change an asset's valuation, because value is a multiple of income. Data is no longer a back-office nicety; it is a direct input to the number the whole business is measured on.

• The fourth is the tenant, who now expects a responsive, digital, self-service experience that simply cannot be delivered without a live operational data foundation underneath it.

None of this is theoretical anymore. Capital is flowing toward the thesis, with roughly $16.7 billion invested in property technology globally in 2025 by industry estimates, and AI adoption among owners and operators climbing from a rounding error a few years ago to the majority of the market piloting tools today. The largest players are reframing themselves openly. CBRE increasingly emphasizes its proprietary data, AI capabilities, and technology platform as key competitive advantages - CEO Bob Sulentic has stated publicly that CBRE holds more real estate data than any company in the world, framing that data as central to how the firm positions itself against AI-driven disruption.

And then there is the tell that should stop any property executive cold. One of the world's most valuable commercial real estate information companies, CoStar Group, owns essentially no buildings. Its core asset is a real-estate database more than thirty years deep, updated tens of thousands of times a day. A significant part of CoStar's valuation reflects the proprietary information, analytics, and marketplace assets it has built - its subscription revenue, analytics tools, marketplaces, and network effects - rather than ownership of physical real estate. CoStar is genuinely an information company rather than a property owner, but its value comes from the combination of data, analytics, and software built on top of that data, not the raw database alone. That combination is the migration, made visible.

Data Is an Asset - and Property, of All Industries, Should Already Know It

Here the argument turns from trend to something more pointed, because "data as an asset" is not a metaphor. It is a claim that survives scrutiny.

The person who made that case most rigorously is Doug Laney, whose work on infonomics - the economics of information as an asset - shows that data satisfies every formal criterion of an asset. It can be owned and controlled. It can be exchanged for value. It generates probable future economic benefit. It meets the definition as squarely as a warehouse does. The only reason it does not appear on the balance sheet is that accounting standards have never caught up - internally generated data, like most internally built intangibles, simply isn't recognized under the rules. Laney's most cutting observation is that because of this blind spot, most organizations manage their office furniture with more asset discipline than they apply to their data.

Now sit with that line inside a property context, because it produces a genuine paradox.

Property is the most asset-disciplined industry on earth. It maintains asset registers down to the fixture. It runs depreciation schedules, capital expenditure plans, condition surveys, reserve studies, title records, and formal valuations. It applies rigor to concrete, steel, and roofing membranes that most industries never apply to anything they own. And it applies almost none of that rigor to the one asset it holds that actually appreciates.

The building depreciates. The data appreciates. The caveat matters, so state it plainly: raw data does not appreciate on its own - only data that is well-governed, complete, and trusted does. But when that condition is met, the direction of travel genuinely reverses. A roof has a known, declining useful life; every year it is worth a little less and moves a little closer to replacement. A decade of clean, continuous operational history moves the other way - every year it captures more, prices the next acquisition more sharply, underwrites the next renovation more confidently, and trains the next model more accurately. The balance sheet tracks the depreciating asset to the dollar and the appreciating one not at all. An entire industry built on asset management is quietly mismanaging its best-performing asset class, largely because that asset class is invisible to its accounting.

Run Your Data the Way You Already Run Your Buildings

The encouraging part of this diagnosis is that property does not need a new discipline to fix it. It already owns a mature, battle-tested discipline for managing assets. It just needs to point that discipline at data. Four transfers do most of the work.

Keep a register. No competent operator manages a portfolio without knowing what it owns and where. Yet most operators cannot say what operational data they hold, which system it lives in, or who controls it. There is no inventory, and without an inventory there is no asset - only exhaust. The first act of treating data as an asset is the same as the first act of treating property as one: take stock of it.

Hold custody. An asset you do not control is not really yours. This is the transfer property leaders most often miss. If leasing, maintenance, accounting, and screening data each live inside a different vendor's system, on that vendor's terms, then the operator has become an operational data company that does not hold title to its own main asset. Deere's own customers learned the harder edge of this: the platform that holds the data holds the leverage. Custody is not a technicality; it is ownership.

Maintain condition and continuity. Physical assets are maintained on a schedule; data needs the same care - captured consistently across buildings and years, kept clean, and not silently reset every time a system is swapped or an experienced manager walks out the door with context that was never written down. History has to be continuous to be worth anything, a standard we examine closely in your AI is only as smart as your data history.

Let it show up in enterprise value. The reason to do any of this is not tidiness. It is that in a data economy, the operational data foundation is increasingly what a sophisticated buyer, lender, or capital partner is actually pricing when they look at the business. Well-kept data is not overhead. It is stored enterprise value.

The Most Expensive Way to Become a Data Company

Most property companies are becoming data companies right now, by accident, and by the most expensive route available: tool sprawl. A point solution for leasing, another for maintenance, another for screening, another for accounting, another for payments - five, ten, fifteen systems, each capturing a slice of the operational record and each holding that slice on its own terms. The operator ends up as a data company whose principal asset is scattered across a dozen landlords it does not control, in formats that were never designed to add up to one coherent record.

This is the custody failure and the continuity failure happening at the same time, and it is the default outcome rather than a rare one. It is also where the enterprise-architecture decision quietly becomes a strategic one, because the alternative is not more integration effort layered on top of fragmentation. It is deciding, deliberately, that the data is yours and architecting the business so that the operational record accumulates in one place you control, rather than pooling inside whichever vendors you happened to buy.

The point worth taking is not about any single product. It is that the operators who consolidate their operational data into an asset they own are the ones positioned to compound it, while the operators running fifteen disconnected tools are compounding someone else's.

The Choice Hiding Inside the Inevitability

You do not get to opt out of becoming an information-driven property company. That decision has already been made for the industry by AI, by capital, by margin, and by tenants. The only decision left is whether to be a solvent one.

The company that denies the transition still generates every byte of the same data. It simply lets that data evaporate through the phone calls no one logs, the context that leaves with departing staff, the records overwritten on a short retention cycle, the fifteen vendors quietly accumulating what should have been the operator's own asset. And it tends to discover the cost at exactly the wrong moment - at a sale, an audit, or a capital raise, when someone asks it to prove its own numbers and it cannot reconstruct the history to do so.

The asymmetry is what makes this urgent rather than merely interesting. Intelligence is converging; increasingly capable models are becoming widely accessible to every operator. Data is not converging. The operator that began treating its operational record as an asset years earlier holds something a latecomer cannot purchase at any price - it can only be grown, slowly, in real time, one captured event at a time. That is the rare kind of moat in a commoditizing world: an advantage that money cannot shortcut.

So the honest version of Porter's question is the one every property leader should be asking now, while the answer is still theirs to shape rather than theirs to discover under pressure. What business am I actually in? The operator who asks it early, and answers it honestly, gets to build the data asset on purpose - with a register, with custody, with continuity, and with the same discipline it already brings to every roof and boiler it owns. The operator who never asks becomes a data company anyway. Just one that spent years throwing its most valuable asset in the bin, and called that normal.

Frequently Asked Questions

Q1. What does it mean that every property company becomes a data company?
It means competitive advantage in property is shifting from the physical asset to the operational data about it - how it performs, what it costs, and what happens next. As AI, capital, and margin pressure reward data capability, the accumulated operational record becomes the differentiator while the building itself becomes a more commoditized input. Every operator generates this data by running the business; the only choice is whether they capture and own it well.

Q2. Is "data as an asset" a real concept or just a metaphor?
It is a real and defensible claim. Data meets every formal criterion of an asset - it can be owned, controlled, exchanged for value, and generates probable future economic benefit - it just isn't recognized on the balance sheet because accounting standards haven't caught up to internally generated data. Doug Laney's work on infonomics makes this case in detail, noting most companies apply less discipline to their data than to their office furniture.

Q3. If we already own buildings, why does data matter more?
Because the two assets move in opposite directions: a physical asset depreciates on a known schedule, while a clean, well-governed, continuous operational history appreciates, pricing deals more sharply and training AI more accurately each year. In a data economy, that operational foundation is increasingly what buyers, lenders, and partners are actually pricing when they value the business.

Q4. What is the difference between a good and a bad property data company?
A good one applies the same asset discipline to data that property already applies to buildings: keeping a register, retaining custody, maintaining continuity, and letting it show up in enterprise value. A bad one generates identical data but lets it evaporate through unlogged events, lost institutional knowledge, and fragmentation across disconnected vendor tools.

Q5. Doesn't buying more PropTech tools make us a stronger data company?
Not automatically, and often the opposite. The common failure mode is tool sprawl - many point solutions each capturing a slice of the operational record on the vendor's own terms, leaving the operator's main asset scattered and hard to recombine. Consolidating data into one owned foundation, not adding more disconnected tools, is what actually turns it into an asset.

Q6. Who actually owns our operational data?
That depends entirely on your architecture. If your leasing, maintenance, accounting, and screening data each live inside separate vendor systems, effective control of your most valuable asset may sit with those vendors rather than with you. Custody - the ability to hold, move, and build on your own data - is what decides whether you're a genuine data company or just a data source for someone else's.

Q7. We're a small or mid-sized operator. Does this apply to us?
Yes, arguably more so, because starting early is the one advantage a latecomer can't buy back. Data can only be accumulated in real time, so a competitor who began capturing clean, continuous history years earlier holds a lead money can't shortcut. Portfolio scale matters far less than the discipline of the record.