Read a week of dance music news and you get a list of endings. A Sheffield club shutting after its final weekend. Several Birmingham venues in the path of a railway expansion. Westminster consulting on rules that would make new music venues harder to open. Ibiza moving to stop issuing licences for new clubs. A Luzern company behind three venues declaring bankruptcy. A Rome community space evicted after a fire. Each of those is real and each of them matters to the people involved. Together they tell a story about landmarks disappearing.
We index event listings across a set of cities, so we can ask a slightly different question. Not which rooms are closing, but where a city's calendar actually sits. The answer surprised us, and it suggests the closure story is being told about the wrong buildings.
What we counted
Every public event our index returned for eight cities over the 46 days from 17 August 2026, which came to 11,935 listings across 3,610 distinct venues. No sampling. The full dataset is published as JSON with its methodology and its limits attached, and the honest headline limit is that this measures our coverage of a city rather than the city itself.
The big rooms carry less than you would guess
| City | Venues programming | Events | Share held by the ten busiest venues | Events per venue |
|---|---|---|---|---|
| London | 646 | 1,990 | 11.6% | 3.1 |
| Berlin | 558 | 1,916 | 13.7% | 3.4 |
| Tokyo | 604 | 1,397 | 16.0% | 2.3 |
| New York | 681 | 2,835 | 16.4% | 4.2 |
| Lisbon | 252 | 556 | 23.9% | 2.2 |
| Los Angeles | 510 | 1,954 | 27.4% | 3.8 |
| Paris | 193 | 479 | 33.8% | 2.5 |
| São Paulo | 166 | 808 | 42.3% | 4.9 |
In London, the ten busiest venues in our index account for under an eighth of everything scheduled. In Berlin it is under a seventh. The busiest single room in either city holds around two percent of its city's calendar.
That is not what the discourse implies. The way club culture gets discussed, you would expect a handful of institutions to carry the scene, with everything else orbiting them. In the cities with the deepest listings, the opposite is true. The calendar is spread across hundreds of rooms, most of which host a handful of things across six weeks.
The median venue is very small indeed
Divide events by venues and London comes out at 3.1 events per room across the window, Berlin at 3.4, Tokyo at 2.3. Those averages are dragged upward by the busy end, so the typical room is quieter still. A venue hosting two or three events in six weeks is not a nightclub in the sense the word usually carries. It is a bar with a booth on Fridays, a gallery that opens late once a month, a restaurant that clears tables at eleven, a rehearsal room that occasionally sells tickets.
This is where the culture actually happens, numerically speaking. It is also the part of the ecosystem that no campaign is organised around. When a named institution is threatened, there is a petition, a press cycle and a hashtag. When a bar quietly stops doing its Thursday because the licence renewal got complicated, nothing happens at all, and the calendar loses a room nobody outside forty people had heard of.
Concentration is a real risk, and it varies enormously
The second thing the table shows is how differently cities are structured. São Paulo's ten busiest venues carry 42 percent of its calendar and Paris's carry 34 percent, against London's 12 percent. Some of that is our coverage rather than the city, and we say so plainly. Some of it is genuine, and it is visible in who those venues are: in São Paulo the top of the list is dominated by a single cultural institution's network of houses.
A concentrated calendar is more efficient and more fragile. When ten rooms hold four in ten events, losing two of them removes a chunk of a city's cultural supply in one go. When ten rooms hold one in eight, the same loss hurts the people involved and barely moves the total.
So the honest reading of a closure headline depends on which kind of city it happened in, and almost nobody makes that distinction when reporting one.
What this changes if you programme events
Three things follow from a long-tail calendar, and all of them cut against how promoters usually think.
- The room you want probably is not a venue. In a city with six hundred programming spaces, most of the interesting slots are in places that do not describe themselves as clubs. Finding those is a research problem rather than a budget problem.
- Competition is not where you think. You are not competing with the landmark down the road for an audience. You are competing with two hundred small things happening the same night, most of which your crowd has never heard of either.
- A room that hosts three events in six weeks has gaps. That is a much better first email than a venue with a full calendar and a booker who is not reading.
The part we cannot measure
A listings index sees what is scheduled, so it cannot see a venue that closed last year or a night that stopped being programmed before we started counting. It measures supply that exists, not supply that used to. If rooms are disappearing steadily, our snapshot shows the survivors and calls it a healthy long tail.
That is a real limit and it points at the study worth doing next, which is the same count repeated at intervals. A single snapshot can tell you the calendar is spread thin across hundreds of rooms. Only a series can tell you whether the number of rooms is falling.