A perfect district that empties after six. What that costs: and what the fix looks like in three places on the Zuidas.
A perfect district that empties after six. What that costs, and what the fix looks like in three places on the Zuidas.
Amsterdam built one of Europe's most successful business districts.
9.4 to 1.
That is how many people are in the Zuidas during the day for every one who is still there at night. An illustrative ratio, built from roughly 47,000 people at the daytime peak and around 5,000 after six.
It is the most important number about the Zuidas that nobody priced.
The Zuidas is not a failure on paper. On paper it is a triumph.
50,000+ people work there. Just over 5,000 live there. Around 32,000 students from the VU and the UvA move through it. Roughly 2,000 companies. 245 hectares. The best public transport in the country. Architecture that wins international awards.
Everything you can draw has been drawn. Everything you can build has been built.
And still. Walk through it at half past seven in the evening and you are standing in a district that is locked.
Picture a party thrown by architects.
The invitations went out. The tent is up. The decorations are world class. Every detail is correct.
Nobody stayed.
A party is the people who stay. And people stay only when there is a reason. A place to eat. A reason to walk. Something to come back to when the laptop closes.
The Zuidas did not fail at architecture. It succeeded at architecture. And it failed at everything architecture is supposed to serve.
One question decides whether a district lives: do people stay after six. That question was never priced. Experience in real estate is not a line on the budget. It counts as atmosphere, as a side effect, as something that will sort itself out.
It did not sort itself out.
The interesting part is that none of this requires tearing the Zuidas down. The bones are good. What is missing is the lived layer, and the lived layer can be designed and operated like any other system.
Three places make the point. Each one is the same failure, and each one has the same kind of fix. Not a render. A system that keeps the place alive after the architect has left.
Valley is a brilliant object. 75,000 square metres, three towers, 200 apartments, 220 plant species, an atrium people travel to see. It also keeps the city at arm's length. The pool, the gym and the concierge are resident-only. The sky bar is reached through a furniture store on Beethovenstraat. The ground floor is something you pass, not something you stay on.
A plinth is the part of a building the street actually touches. It decides whether a frontage is a wall or an invitation.
The change is not architectural. The towers stay. What changes is who the ground floor is for, and whether someone is responsible for keeping it alive on a Tuesday in February.
George Gershwinplein is a square in name. In use it behaves like a forecourt: a space you cross on the way to somewhere, not a place you arrive at. The geometry says square. The experience says transit.
A forecourt and a neighbourhood square are made of almost the same materials. The difference is intent, and a system that maintains the intent after opening day.
The "after" is not just a nicer picture. It is a buurtplein plus the system that runs it. Squares do not stay alive on their own. Someone has to own the calendar, the upkeep and the standard, across operator changes and across years. Design makes the square possible. The operating layer makes it last.
Zuidplein has a split personality. On a weekday it works. Lunch crowds, meetings, movement. On a weekend it empties, because everything around it was programmed for the office and nothing was programmed for the people who might otherwise stay.
A place that is only good five days out of seven is not a place. It is a schedule.
The fix is to design for the weekend with the same seriousness the weekday already gets. Same square, two clocks, one standard.
This is the pattern under all three.
Experience is not soft. It is causal.
Real estate and hospitality treat experience as an unpriced externality. A soft cost that loses every trade-off against hard euros. A building is measured in square metres, in monthly rent, in yield. Whether anyone actually wants to be there appears in no model.
The plan was not blind to this, to be fair. Around 2,774 homes have been built so far, and another 7,000 or so are planned through 2030, with a mandated mix of 40 percent social rent, 40 percent mid-market and 20 percent upmarket. The ambition is on paper in black and white. The office side is all but complete. The lived side trails by years. The ambition was right. The order and the pace were wrong.
That is a structural failure. And the market is starting to reprice it.
The three places above show what repricing looks like in practice. Not demolition. A plinth that invites. A square that holds people. A place that works on Sunday. Each is a design decision followed by an operating commitment, which is the part real estate keeps forgetting to fund.
The Zuidas is the proof at city scale. 245 hectares of evidence that you can build a perfect district and still forget what a district is for.
This is episode 1 of six. The coming parts run through the residents who say it themselves, the plinth in close-up, the maths of building for capital, and Valley as the Zuidas in miniature. It ends on the question that holds the whole thing together: what changes when experience becomes a line in the pro forma.
For now, one line to keep.
A party is the people who stay.
Figures: zuidas.nl (2024). The 9.4 to 1 ratio and the day-versus-night presence numbers are illustrative, derived from daytime versus evening presence, not an official statistic. Valley details per MVRDV and public reporting. The three before/after visuals are design propositions, not approved plans.
Episode 1 of Experience as infrastructure. The Dutch long-read follows on Substack.