PropTech Connect and Mitsubishi Electric just published a research report worth reading: Bridging the Landlord-Tenant Divide with Smart Building Technology. It's built on a July 2026 roundtable with CBRE, British Land, Legal & General, M&G Real Estate, CapitaLand, Hines, Tishman Speyer, WSP, Cushman & Wakefield, Avison Young, and Urbanest. Real owners and operators, not vendors pitching a platform.

Its central argument lands cleanly: a building isn't smart because it has sensors, an app, or a dashboard. It's smart when information changes a decision and improves an outcome. Most of the industry is still stuck at the first part.
‍

The stat that says more than the framework

Buried in the report is a CBRE data point that does more work than any of its frameworks: average global office utilization reached 53% in 2025, while peak utilization averaged 80%. That's not a small gap. It means most buildings are being run for an average occupancy that barely resembles how the space actually gets used on its busiest days. Heating, cooling, cleaning, staffing, all of it sized to a number that mostly doesn't happen.
‍

The report's answer is operational: move from fixed schedules to demand-based control, let occupancy data adjust HVAC, lighting, and staffing in real time. That's the right fix for a building that already exists. But it treats the 53/80 gap as something to manage after the fact, once the building is built, leased, and occupied.
‍

The gap usually starts earlier than that

Most of the time, that mismatch traces back further than the building's controls. It's a planning problem, decided long before anyone installs a sensor. Nobody designed the space around how it would actually flex. How many people show up on a Tuesday versus a Thursday, how much space a team of 40 really needs when a third of them are hybrid, what happens to a floor plate when headcount changes mid-lease. That gets decided once, early, often without much data at all. Everything downstream, the BMS, the sensors, the "responsive" building, is left compensating for a decision that was never built to flex.
‍

The report itself gestures at this without naming it. Its Hines contributor puts it plainly: "The earlier you can partner, the better." Its CapitaLand example, where the owner funded metering and upgrades in exchange for a longer lease term, worked because the investment was tied to a real asset-management decision, not bolted on afterward. Retrofitting intelligence into a building that was planned blind is expensive and slow. Planning with the data from the start isn't.
‍

Why the split incentive keeps showing up

The report is honest about the other barrier: landlords fund the technology, tenants get the lower bills, and neither side wants to move first. That's a real problem, and no amount of smart building tech resolves it on its own. It needs a data charter, shared savings, or a lease-event trigger, exactly as the report lays out.
‍

This incentive problem gets easier to solve the earlier it's addressed. A landlord who can show a prospective tenant exactly how a space will perform, not just what it will look like, has a much easier conversation about who pays for what, because the value is visible before the lease is even signed.


The layer before the smart building

None of this is a knock on the report. If anything, it's the natural next question it raises. Smart building technology makes an occupied space responsive. Space intelligence makes sure the space was planned to be responsive to begin with. One picks up where the other should have started.
‍

We made a version of this argument recently from the other side of the lease: in Office Vacancy Is Falling Faster Than It Has in a Decade, we wrote about how, in a tightening market, the bottleneck isn't the building. It comes down to how fast an owner can show a tenant what an empty space could become. This report picks up the story just after the lease is signed. Put the two together, and the pattern is the same at every stage: the winners are the ones who use data to make the decision, not just to describe the space after the decision's already been made.

If you're planning space, not just operating it, that's the layer worth thinking about.

‍

Ronit is CPTO at qbiq, where architect-verified space plans ship in a day, not a week. qbiq's Agentic SpacePlan is now in early access for architects.