The 10% Office
The consensus narrative on AI and office space runs like this: AI makes workers more productive, productivity means fewer workers, fewer workers means less space. Sell office, buy industrial.
I think this misses what’s actually happening.
I’ve spent the past year rebuilding my own practice around AI tools. The compression is real. Work that took days now takes hours. Work that took hours now takes minutes. Reporting that consumed 80% of my time now consumes perhaps 10%. The productivity gain isn’t marginal. It’s structural.
But here’s what I’ve noticed: I didn’t stop needing space to work. What changed is what I use space for.
The Compression Reality
Let’s be specific about what AI compresses. It’s the busy work. The drafting, the formatting, the information gathering, the first-pass analysis, the routine cognitive tasks that filled calendars and gave the appearance of productivity. This is work that could be done anywhere, and increasingly, can be done by machines.
What doesn’t compress is judgment. Synthesis. The moment where you look at all the information and decide what it means. The conversation where trust gets built or broken. The room where the decision gets made.
This is the 10% that remains. And this 10% has different spatial requirements than the 90% that’s disappearing.
Beyond the Fit-Out Fallacy
Here’s where most analysis goes wrong. The instinct is to say: make the office nicer. Add breakout zones. Install a coffee bar. Paint it bright colours. Create “collaboration spaces.”
This misses the point entirely.
A fancy fit-out means nothing if it sits inside a dysfunctional culture. I’ve walked through beautifully designed offices where people still retreat to their desks and put headphones on. The space signals collaboration. The behaviour signals isolation. The fit-out is theatre.
The 10% that remains after AI compression isn’t about aesthetics. It’s about genuine human connection. The moments where presence matters. Where reading the room changes the outcome. Where trust forms because you showed up, not because you logged on.
No amount of designer furniture creates that. Culture creates that. Space can support it or obstruct it, but space alone doesn’t generate it.
The Investor’s Dilemma
This is where it gets interesting for capital allocators.
If you’re an office investor, you don’t control tenant culture. You don’t control how organisations choose to work. You control the base building and the common areas. That’s it.
So the question becomes: what can you do with what you actually control?
Every tenant in a conventional office building faces the same problem. They need space for the 10%. They need rooms for the conversations that matter, the synthesis sessions, the relationship moments. And every tenant solves this problem independently, inside their own demise. They each build their own meeting rooms, their own collaboration zones, their own “third spaces.”
The result is duplication and mediocrity. Every tenancy has meeting rooms that are too small, booked solid, and acoustically compromised. Every tenancy has a breakout area that looks good in the marketing photos but sits empty most of the time. Every tenant pays for space they underutilise while lacking space they actually need.
The Third Space Opportunity
What if the building provided the third space instead?
Not as an afterthought. Not as a token lounge next to the lobby. As genuine infrastructure for the 10%.
Think about what tenants actually need for the work that remains:
Space for six people to work through a problem for four hours
A room where a difficult conversation can happen privately
Somewhere to host clients that isn’t a cramped meeting room
A place for the whole team to gather quarterly
Environments that vary in energy: quiet focus, active collaboration, social connection
No single tenancy can economically provide all of this. But a building can.
The investor who controls the base build and common areas has an opportunity to create shared collaboration infrastructure that tenants tap into as needed. Space that no individual tenant could justify, but that collectively serves everyone.
This changes the value proposition. The tenant doesn’t need to build their own collaboration zones. They take less demised space and access shared infrastructure for the moments that matter. The building becomes a platform, not just a container.
What This Actually Looks Like
I’m not talking about a nicer lobby or a rooftop bar. I’m talking about genuinely useful space:
Bookable rooms at various scales, properly soundproofed, with technology that actually works. Not meeting rooms as an afterthought, but meeting rooms as core infrastructure.
Flexible event space that a tenant can use for a quarterly all-hands without paying for external venues.
Quiet zones for focused work when home isn’t working and the open plan is too loud.
Informal spaces designed for the spontaneous conversations that video calls can’t replicate.
The key is that this space must be genuinely good. Not gesture space. Not marketing space. Space that solves real problems better than tenants can solve them independently.
The Bifurcation, Revisited
This reframes the market bifurcation.
The question isn’t just “premium versus commodity.” It’s whether a building offers infrastructure for the 10%, or just square metres for the 90% that’s disappearing.
Buildings that figure this out have a value proposition that survives the compression. They’re not selling space. They’re selling capability. The capability to do the work that still requires presence.
Buildings that don’t figure this out are competing on cost for a use case that’s evaporating. They’re selling square metres to tenants who need fewer square metres every year.
What I’m Watching
I’m paying attention to how tenants are using common areas. Not what the marketing says. What actually happens. Are the shared spaces empty, or are they solving real problems?
I’m watching lease structures. Are tenants taking less demised space and relying more on shared infrastructure? That’s the signal that the model is working.
I’m watching the buildings that are experimenting with genuine third space provision. Not the ones adding a coffee machine to the lobby. The ones rethinking what common area actually means in an era where individual task work is disappearing.
And I’m watching culture. The buildings that attract tenants with healthy cultures will outperform. Not because of the fit-out, but because healthy cultures actually use collaborative space. Dysfunctional cultures retreat to desks regardless of what surrounds them.
The Uncomfortable Truth
Here’s what investors don’t want to hear: you can’t control whether this works.
You can build the best third space infrastructure in the market. If your tenants have toxic cultures where no one wants to be in the office, the space sits empty. The building can enable the 10%, but it can’t force it.
This means tenant selection matters more than it used to. The quality of organisations in your building affects the value of shared infrastructure. Buildings that attract organisations with genuine collaborative cultures will see that infrastructure utilised. Buildings that attract organisations in denial about their dysfunction will see beautiful spaces gathering dust.
For Capital Allocators
If you’re holding office, the question isn’t “will demand recover?” It’s “does this asset offer something tenants can’t replicate themselves?”
If you control a building with potential for genuine third space infrastructure, there’s an opportunity to reposition. Not through cosmetic upgrades, but through rethinking what the common areas actually provide.
If you control a building that’s just floor plates with a lobby, the calculus is harder. You’re selling the 90% use case to tenants who need less of it every year.
The market will figure this out eventually. The investors who see it now have a window to act.
What I Don’t Know
I don’t know how fast tenants embrace shared infrastructure over self-contained demises. Organisations are conservative. Lease structures are sticky. The shift may take longer than the logic suggests.
I don’t know which building operators will figure out how to run third space well. It’s a different capability than traditional property management. Some will execute brilliantly. Many will create gesture spaces that solve nothing.
And I don’t know how to value this yet. The market doesn’t have a model for “building as collaboration platform.” The premium for genuine third space infrastructure is theoretical until transactions prove it.
If you’re thinking about this differently, then I’d love to hear it.
Steven McCormack MRICS January 2026