In August, Athens holds 4.2% of its year's nightlife - its quietest month by some distance. In November it holds 11.4%, and in May exactly the same again. The city's night is 2.7× bigger in the month nobody flies in for than in the month everybody does.

That is not a quirk of one city. Thessaloniki does the same thing, and Amsterdam and Rotterdam do the exact opposite - their years fill up in summer and thin out in January. Four cities, two countries, two calendars running against each other.

The short version

Athens is at its quietest in August - 4.2% of its year - and its busiest in Nov and May, at 11.4%. The city's night is 2.7× bigger in November than in the month everyone visits.
The Greek and Dutch nights run on opposite calendars: Greek cities average 1.98× as much activity in winter as in summer, Dutch cities 0.9×.
Amsterdam's single biggest month is October at 14.8% - the month ADE takes over the city.
Greek rooms are club rooms: 55.0% of a Greek city's events are a party or a DJ set, against 40.6% in the Dutch cities.

Four years, four shapes

Each chart is one city's own twelve months, and the bars add to 100% across the row. The dotted line marks 8.3% - where every month would sit if a city's night were evenly spread. All four share one vertical scale, so the shapes are read against each other.

Athens peak Nov & May 11.4% · trough Aug 4.2% 2.09× winter

Thessaloniki peak Nov 12.6% · trough Aug 5.1% 1.86× winter

Rotterdam peak Jul 12.8% · trough Aug 4.9% 0.98× winter

Amsterdam peak Oct 14.8% · trough Jan 5.8% 0.81× winter

The month Athens leaves

August is the single quietest month in both Greek cities - Athens 4.2%, Thessaloniki 5.1% - and the reason is not that Greeks stop going out. The city empties: the night moves to the coast and the islands, and the rooms that stay open are seasonal ones on the water. Our own price work found the same thing from the other end - the Athens summer clubs that publish a door price sit on the coast road, not in the city.

Two calendars

One number separates them: how much of a city's year lands in November and December against August and July. Above 1× is a winter city; below, a summer one.

Athens 2.09×

Thessaloniki 1.86×

Rotterdam 0.98×

Amsterdam 0.81×

The split is clean and it runs by country, not by size: both Greek cities are winter cities, both Dutch cities are summer cities. It is worth saying what this does not show - two countries is not a north/south law. It is four cities, named, on one method.

Amsterdam in October

Amsterdam's biggest month is not a summer one. It is October, at 14.8% of its year - nearly double the flat line, and the sharpest single spike in this report. October is when Amsterdam Dance Event takes over the city.

A number we did not set out to find

ADE reports 600,000 visitors across 1,200+ events - an organiser's figure, which our report № 002 cites as one. This chart was built from event dates alone, with no knowledge of ADE, and the spike is there anyway: roughly 800 events above Amsterdam's own baseline month. Two independent counts of the same week, and they agree in direction and rough size. We hold the month and not the day, so read this as consistent with ADE rather than as a measurement of it.

Club city, live city

The same data answers a different question: of everything a city puts on, how much of it is a dancefloor? Parties and DJ sets against live sets and concerts, each as a share of that city's own events.

Athens 55.4% 29.9%

Thessaloniki 54.6% 32.5%

Rotterdam 39.8% 35.5%

Amsterdam 41.3% 31.1%

Greek cities run about 55.0% club against the Dutch 40.6%. That is the same fact that defeated our price work from a different angle: a city whose nights are mostly dancefloors is a city whose entry prices are settled at the door and never published.

Which cities can be measured

Four, of the twelve in the catalogue. A seasonal shape only means something if the whole year is actually there, so a city qualifies only when all twelve months carry real volume. Eight do not, because we began indexing them partway through the window - and their apparent seasonality is spectacular for exactly that reason. One excluded city appears to be 5,688× busier in its peak month than its quietest, because its quietest month contains one event. That is a picture of our crawler, not of a city, and it is why those eight are named here rather than quietly left out.

Timișoara
Bucharest
Paris
London
Berlin
Barcelona
Patras
Cluj-Napoca

The same discipline governs everything above: every figure in this report is a share of one city's own year. We index Athens far more deeply than Berlin, so any count set against another city's count would measure our coverage rather than their nightlife. There is no "busiest city" here and there should never be one.

Method & reuse

The data. Every live, publicly listed event in Mood's own catalogue with a local date between August 2025 and July 2026 - 55,611 of them, drafts and hidden records excluded. Each event contributes its month, its city and its format. No ticket or price field is involved anywhere in this report.
Why these are shares and never counts. Our indexing depth varies enormously by city, so a table of event totals would rank our crawler rather than Europe's nightlife. Every figure here is therefore a proportion of a single city's own twelve months, which is unaffected by how deeply that city is indexed. The underlying counts exist in the data file; they are not a finding.
The twelve-month guard. A city is published only if each of the twelve months carries at least 100 events. This is the guard that separates a quiet month from an absent one, and without it eight cities would produce dramatic seasonal claims that are pure ingestion history. Excluded cities are named in § 05.
What one year cannot tell you. This is a single cycle, so season and trend cannot be separated: a city that grew over the year would look like a city that peaks late. The Greek and Dutch shapes are strong enough, and opposed enough, that growth alone is an unlikely explanation - but a second year is what would settle it, and we do not have one yet.
Formats. "Club" is party and DJ set; "live" is live and concert. Festivals, panigiria and other formats are counted in each city's total but sit in neither share, which is why the two do not add to 100%.
Reuse. Data as CSV, CC BY 4.0. Cite as: Mood Data Desk (2026). When Europe Goes Out. Mood - Music of Our Desire. https://www.musicofourdesire.com/reports/when-europe-goes-out