Data Centers Could Learn a Lot From the Best Airports in the World

There's a conversation happening on social media right now about making data centers look better. Architects, venture capitalists, and armchair designers are generating AI renders of hobbit-hole data centers, medieval castle data centers, Parthenon-inspired data centers. Some of it is genuinely beautiful. And, I think they're onto something.

Better looking data centers would be a massive positive. For communities. For the companies building them. For the broader public conversation about AI infrastructure. The instinct is right.

I just think the conversation is stopping too short.

I spent the better part of two decades selling experience design services, mostly in the digital space. UX, product design, the whole ecosystem. What that work actually taught me is something that applies just as much to a building as it does to an app: experience isn't just about how something looks. It's about how it feels to everyone who has to live with it. And when you get that right, it becomes one of the most powerful selling tools you have.

Think about airports.

Airports had every reason to be terrible. Massive infrastructure. Rigid security requirements. Enormous commercial pressure. An endless parade of stressed, exhausted, often frustrated people moving through them around the clock. And the best ones in the world, Changi in Singapore, Hamad in Qatar, even the better terminals at LAX and JFK, figured out that investing in experience wasn't a luxury. It was a competitive and political strategy. Natural light. Indoor gardens. Art installations. Acoustic design. The physical environment communicates something to every person who passes through: we thought about you.

That's not a small thing. That's a sale. Every time someone walks through and feels something other than dread, the airport is making a case for its own existence.

Hospitals have learned some of this too, though more slowly. There's real research showing that access to natural light, proper air circulation, and thoughtful physical design affects patient recovery. Some of the best hospital systems have started treating the environment as part of the treatment. Federal buildings and civic architecture have gone through similar evolutions, sometimes for the better, often still frustratingly not. But the through line in every case where it worked is the same: someone made the argument that the investment in experience would pay for itself in goodwill, in faster approvals, in community relationships that don't have to be managed through crisis.

Data center builders haven't made that argument. Yet.

Right now, data centers feel imposed. They arrive in communities looking like what they are: a massive commercial operation optimized entirely for the people running it, with zero consideration for the people living next to it. Windowless concrete boxes. Hundreds of acres of fencing. And increasingly, the hum and roar of generators and cooling systems loud enough to register as a genuine quality-of-life complaint.

That noise piece doesn't get nearly enough attention. The aesthetics conversation is mostly about what these buildings look like from the outside. But residents near data centers are filing noise complaints, attending town meetings, and in some cases successfully blocking permits, not because of how the buildings look, but because of how they sound. Constant low-frequency hum. Generator tests at odd hours. HVAC systems running at full capacity around the clock.

This is a solvable engineering problem. Quieter machines exist. Acoustic dampening exists. The technology to monitor and actively manage noise output in real time exists.

The question is whether anyone is building it into the business case.

My guess? They're not. And I think I know why.

An architect quoted in the Fast Company piece that sparked this conversation said it plainly: "No underwriting model has a line item for vibes."

He's right. And he's also describing a gap that smart operators should be racing to fill.

Because here's what I've seen over and over again in experience design: the projects that invest in how something feels to the people around it consistently spend less on community relations, permitting battles, and reputation repair over the life of the asset.

The investment in experience isn't a cost. It's risk mitigation. It's faster time to approval. It's the difference between a community that tolerates your facility and one that actually points to it with some degree of pride.

And when you're trying to build 3,000 more data centers in a country where communities are increasingly organized and permitting is increasingly contested, that difference is worth real money.

This is a sales and business development problem as much as it is a design problem.

The operators who learn to sell the experience of a data center, to make the case to a community before the bulldozers show up, are going to have a meaningful advantage over the ones still treating community relations as a legal and PR function. The ones who invest in experience design are going to close faster, face fewer delays, and spend less on the back-end friction that kills project timelines.

Other countries have figured parts of this out. There are data centers in Europe and Asia that blend into their surroundings, that use architecture as a genuine strategy for community acceptance. It works. It just hasn't scaled here, probably because cost, speed to market, and a culture inside the industry that has historically had to answer to almost no one made the investment easy to skip.

That's changing. The communities getting the next wave of these facilities are going to have more leverage than the ones that got the first wave.

There's an inside-out version of this problem too, and it's where some of the most interesting work is happening right now.

Part of what I do involves working with technology companies building the infrastructure layer inside data centers. One of them, Vectis by Metaspex, is developing software called Vectis that operates at the edge of AI data center environments, making autonomous decisions about power management, thermal control, and fault prevention in real time. The premise is that a smarter building, one that reacts faster and fails less often, consumes less energy and creates fewer of the runaway cascades that make data centers so resource-intensive in the first place.

Better engineering from the inside can meaningfully reduce the environmental footprint that communities are reacting to from the outside.

It's a different lens on the same problem. If the outside-in answer is architecture and experience design, the inside-out answer is software that makes these facilities genuinely more efficient and less damaging to the environment around them. Both matter. Both require someone making the argument that the investment is worth it.

So here's what I keep sitting with: is the reason data centers look and sound the way they do primarily a cost problem? A speed problem? A culture problem inside an industry that grew up without having to care?

I think it's some combination of all three. But I also think the calculus is shifting faster than most operators realize. The communities getting the next round of these facilities are paying close attention to what happened to the communities that got the first round.

The best airports in the world didn't get great by accident. Someone had to make the business case that experience was worth investing in. Someone had to sell it internally before the architects could do anything with it.

That's the conversation I think the data center industry needs to be having.

What do you think is actually holding it back?