How’s Demand? A Better Answer to “How’s the Market?”

By Matt Osowski, SIOR, Executive Vice President, NAI Ohio Equities

It happens at every SIOR meeting, client lunch, and chance meeting in a hotel lobby. Someone asks, “How’s the market?” Most of us answer the same way. We talk about vacancy, net absorption, how many square feet are under construction, and whether rents are rising or flattening.

Those are good numbers. But peel back the onion and you’ll see they aren’t what drives the market. They are what the market leaves behind.

In industrial real estate, nothing gets built, nothing gets sold, and nothing gets leased unless users need space. Developers break ground because they expect occupiers to show up. Investors set prices based on who will fill the building. Vacancy only drops when a user signs a lease. Every metric we quote comes from one thing: a company deciding it needs space.

So how are we measuring that?

The Generality Problem

Ask a broker “How’s user demand?” and you’ll hear something like: “It’s good. It’s not bad. Tenants are active. We’re seeing a lot of things moving around.”

I’ve said it myself. We quote vacancy to a tenth of a percent, then describe demand, the thing that actually matters, as a feeling.

Part of the reason is that demand is hard to pin down. An occupier’s choice to expand, consolidate, relocate, or wait depends on many forces at once: cost of capital, freight costs, fuel costs, labor costs, construction costs, inflation, and consumer confidence. On top of those come political forces (tariffs, trade policy, interest rate decisions) and social ones (e-commerce, reshoring, workforce expectations). Every one of them affects whether a user signs, and they rarely all point the same way.

No wonder we fall back on “Tenants are active.”

What We’re All Using

Before proposing anything new, it’s worth taking an honest inventory of the vocabulary we already share. The metrics our industry quotes are good ones, and each tells you something real. What matters is knowing what each one is actually measuring, and when.

Vacancy rate. A snapshot of what is left over after demand has already been satisfied. It describes the balance of supply and demand last quarter, not the decisions being made this quarter.

Net absorption. The closest thing we have to a demand metric, and still a record of signatures. A lease signed in September reflects a decision made twelve to eighteen months earlier.

Construction pipeline and starts. A measure of developer conviction. That is a forecast of user demand rather than evidence of it.

Asking rents and concessions. Price, not volume. Asking rents are among the last things to move, and concessions move first and quietly.

Leasing volume, tour activity, and requirement counts. A genuine early signal, and mostly unpublished. This is the real basis for “tenants are active,” and it lives in brokers’ heads rather than in anyone’s data set.

Sublease availability. One of the better early warnings that occupiers have more space than they need.

Notice what those have in common. Almost all of them are measured inside the building, after the decision has been made. None of them tell you why a company concluded it needed space, or what would make it decide differently next quarter. For that you have to look upstream, at the economy the occupier is actually operating in.

Which raises a question I would genuinely like answered: what are the rest of you using? When a client asks how demand is, what do you reach for, and how do you articulate it?

Gigawatts, Not Dock Doors

Nowhere is the gap between the scoreboard and the forces behind it clearer than in industrial capital markets.

Industrial-adjacent equities, REITs and private real estate funds have had a strong run in 2026. The easy conclusion is that this confirms robust warehouse demand. I don’t think that’s the right read.

A significant amount of that capital is chasing AI and data center buildout, a business with very little in common with a tenant signing a 300,000-square-foot distribution lease. Different customer, building, economics, and unit of measure.

Warehouse demand is tracked in square feet. Data center demand is tracked in megawatts.

By that measure, demand has been extraordinary. Net absorption across North America’s primary data center markets reached 1,456.2 megawatts in the first half of 2026, while vacancy fell to a record-low 1.4%. That is a remarkable demand story, but it is a story about power capacity, not warehouse occupancy.

Confuse one for the other and you can easily misread what’s happening on the ground.

I don’t have to look further than my own backyard to see it.

Licking County, and the New Albany corridor in particular, has become one of the most competitive industrial land markets in the country. Not because distributors suddenly need more dock doors, but because hyperscale users need power and fiber.

Microsoft has assembled more than 700 acres in Licking County, including approximately 183 acres in New Albany purchased for $56.9 million after a nationwide site search. Cologix acquired approximately 154 acres in Johnstown for $45 million and announced plans for an eight-building campus capable of delivering up to 800 megawatts, representing more than $7 billion in potential investment. A Meta infrastructure affiliate also acquired a New Albany parcel for approximately $42 million around the turn of 2026, although Meta had not confirmed the intended use when the transaction was reported.

A decade ago, much of that land would have been evaluated primarily through a distribution lens. Today, some of the most aggressive bidders are measuring value in available power, transmission access and fiber connectivity. Raw land can carry one value as conventional industrial ground and an entirely different value when it has data center entitlements and access to the grid.

That shift doesn’t just affect land values. It changes who shows up in the market.

Our own deal flow reflects it. A disproportionate share of the large-format requirements we’re tracking in the corridor can be tied, directly or indirectly, to data center construction, power infrastructure or colocation providers rather than traditional distribution users.

I don’t have an industry-wide statistic to prove that point, and I’d be skeptical of anyone who claims they do. Big-box leasing activity is not consistently published by tenant type. But the distinction is worth acknowledging because two very different users can both be driving industrial demand while pursuing fundamentally different real estate for fundamentally different reasons.

The capital markets may be telling a very real growth story. The challenge is understanding which growth story they’re telling.

If REIT performance, equity valuations and investment flows are being driven by AI infrastructure, they may tell us a great deal about demand measured in gigawatts and very little about demand measured in dock doors.

And if the scoreboard can be that misleading, the obvious question becomes: What should we be watching instead? Not another lagging metric, but the forces a company is weighing before it ever calls a broker.

The Forces Behind a Signature

Upstream of every lease is a company balancing several economic forces at once, and they rarely point the same direction. Manufacturing output and freight volumes matter most, because they physically move goods and create the need for buildings — when production expands or import volumes shift, a real estate requirement tends to follow within a quarter or two. Everything else finances that activity, prices it, staffs it, or reports on it after the fact. Cost of capital decides whether an expansion pencils. Energy and input costs hit the P&L before they hit the footprint. Labor availability shapes where a requirement can even land. And industrial fundamentals — the scoreboard we already spend most of our time quoting — are useful mainly to confirm what the other forces already told you.

Reading them means reading them as an occupier would, not as the broader economy would. Cheap natural gas helps a manufacturer’s decision; record copper prices are a cost, not a demand signal. Capital markets deserve a warning label of their own: through 2026, industrial equities have been bid up on a story that has nothing to do with distribution.

We eventually built a scoring tool for our own team to track these forces systematically against each other — more on that later — but the discipline matters more than the tool: watch what’s moving goods, what’s financing that movement, and what’s staffing it, and you’re watching demand more honestly than any single quarter’s vacancy rate can show you.

When the Forces and Headlines Disagree

The reason to watch these forces rather than the scoreboard is that the two come apart, sometimes for a year at a time.

By mid-2022 the conditions behind demand had already slid from a reopening peak into neutral territory, while national vacancy was still near record lows. For months the metrics we quote described a landlord’s market that the underlying forces had already left. Tenants and landlords reading the conditions had a far better sense of where leverage was heading than those watching vacancy alone.

September 2026 offered a cleaner version of the same lesson. Labor turned sharply better — August payrolls came in at roughly three times consensus, with the prior two months revised up. Energy turned sharply worse, as retail diesel hit an all-time record of $5.85 a gallon, more than $2.00 above a year earlier. The two nearly canceled each other out on the surface, so the market looked calm at a glance. Underneath, the story was specific: this was not a demand problem, it was a cost problem.

For a plant manager or a CFO, that distinction is the entire conversation. It tells them the urgent line item that month was the transportation budget, not rent — a very different message than “tenants are active.”

Reading the forces also teaches patience. Most quarters over the past decade would have landed in a narrow middle band: sideways is this market’s normal state, not an exception. That is why the direction of the forces matters far more than any single month’s reading.

So What Are You Watching?

That is my answer to the question, and it is only one answer. The part I am genuinely curious about is yours. What are you watching to gauge user demand in your market, and how do you put it into words for a client who wants more than a vacancy rate?

This is where I think SIOR can make a difference. We are the most experienced practitioners in the business, and we sit closest to the decisions that create demand. If members nationwide paired a consistent read on the economic forces with their own local knowledge, we could answer “How’s the market?” with something far better than a number describing last quarter.

And the next time someone asks how the market is doing, try answering a different question first: How’s demand?

P.S. — The scoring tool I mentioned earlier is real. Our team built it to weigh these forces systematically against forty quarters of history, and changing any single weight recalculates the whole series. If you’d like to see what your own assumptions would produce — which forces you’d weight higher, which you’d drop — I’m happy to share it. But the conversation I’m actually after is the one above.

About the Author: