The headline for anyone selling a night in Seoul: the most valuable customers are not the ones the market is built around. Older, higher spending guests and people who come for the music are underserved, and the cheapest wins on offer are safety, daytime formats and genuine music variety, not bigger discounts.

Most of these findings describe friction rather than preference. People are not saying they want less nightlife; they are describing the costs that make them go out less: fear, getting home, noise, sameness. Remove a cost and the behaviour the data hints at (more frequent, longer, better nights) has room to show up.

Read together, the seven findings fit a picture of a night time economy optimised for one archetype (young, drinking, club centred) while its actual participants are more varied. The study is exploratory and cross sectional, so the findings are best treated as hypotheses about segment mismatch, to be tested in later waves.

There is clear demand for programming that is not on every other flyer: house nights, 80s and 90s, disco, Latin. Variety is a positioning opportunity, especially for the 30 to 34 group, who are the most DJ focused audience in the data.

Sameness is a reason to stay home. When every room plays the same set, there is little reason to try somewhere new, and the people who care most about music are the first to stop going out.

The music findings map onto Sarah Thornton’s idea of subcultural capital: taste as a marker of belonging. The K-pop paradox, with expats more likely to choose it than Koreans, suggests that what counts as “club music” depends on cultural position. Genre labels were respondent chosen, so a future wave should test them against listening data.

Each district is a different audience, so location is a targeting decision. Hongdae and Seongsu reach the youngest crowd, Itaewon the widest mix and the oldest guests, Apgujeong and Cheongdam the biggest budgets. A night that fits its district starts with half the work done.

People choose neighbourhoods the way they choose friends: for familiarity. Itaewon pulls harder the older people get, which suggests habit and comfort rather than fresh discovery, while the youngest are building their own map in Hongdae and Seongsu.

The age gradient across districts suggests generational sorting in urban leisure space. With one wave of data it is impossible to tell whether Seongsu is rising or simply young; tracking the same question for three to five years would show whether the centre of the scene is moving.

The demand is specific and largely unmet: daytime parties, night markets, outdoor cinema, sober and early evening formats. These are new products, not new discounts, and the guests who want them most are also the ones most willing to pay.

Many requested formats remove a cost rather than add a thrill: daylight solves the last train, early hours solve the next morning, sober nights solve the pressure to drink, no phone floors solve the feeling of being watched.

Stated interest is not attendance. A vote for sober parties or night markets measures aspiration; the next wave should pair these questions with willingness to pay or a real ticketed test to separate intention from behaviour.

Safety is an unclaimed brand position. Only 21.6% of respondents rank safety among the top three things that make a good venue, yet 44.6% of women have felt unsafe, so the first rooms to make safety visible (charter at the door, trained staff, a clear way to report) can own the idea in their district. Women are also the guests most likely to bring a group, so losing one often means losing several.

The written answers suggest one bad experience reshapes habit: a woman who no longer goes dancing because she expects to be groped is planning around the worst night, not the average one. That is why visible reassurance can matter more than incident statistics. The next wave should ask directly whether people went out less afterward.

The 2.6 times gender gap and the coded incident types support treating harassment and transport as distinct mechanisms within a single safety construct. This survey did not measure satisfaction, so the link between feeling safe and enjoying a night cannot be estimated here. Adding a satisfaction scale and a question on changed behaviour would let the next wave test it directly.

Transport is a product feature hiding in plain sight. A venue or promoter that solves the ride home (a partner taxi code, a shuttle to the night bus, an end time set to the last train) is removing the single most cited reason people leave early or stay home.

Missing the last train turns a night into a decision tree: pay a surge fare, hunt for a taxi that will stop, or stay until five. The after hours numbers suggest many people choose to stay, which means part of the late night economy exists because the subway stops, not because people wanted a longer night.

The age pattern (peaking at 65.2% for 25 to 29 year olds, then falling to 44.4% at 35 to 39) suggests transport failure interacts with life stage and income. This chapter speaks to urban mobility research on the night time city; a future wave could add travel time and home district to separate distance effects from behaviour.

One message will not reach both audiences. Koreans lean to clubs and house and techno, expats to rooftops, cocktail bars and hip hop, and they find nights through different channels. Bilingual promotion is not a translation job; it is two campaigns that share a venue.

The two groups share streets but not habits, which is typical of mixed neighbourhoods: overlap in space, separation in routine. Itaewon is where the groups meet, and also where small differences in expectation (door policy, pricing, music) are most likely to turn into friction.

The pattern reads as parallel consumption communities within one scene, in the sense of Will Straw’s idea of scenes as overlapping networks. The binary Korean or expat category is a limitation: gyopo, long term residents and visitors likely sit between the two, and splitting them out is a priority for the next wave.

Women are the stronger signal for format innovation: more interest in cocktails, non alcoholic options, sober parties and art or cinema nights. Products designed around those needs (seated early sessions, mocktail menus, women led lineups) address the guest who is most likely to leave the market if nothing changes.

The spending gap and the safety gap are probably linked. People who feel less secure tend to plan shorter, more contained nights, which shows up as lower spend. Read the lower spend as caution, not lack of interest.

Gender is the strongest single split in the safety data and a meaningful split in drinks and formats. Because the sample is self selected, the absolute percentages should not be generalised, but the direction and consistency of the differences are robust enough to justify gender as a design variable in any follow up.

The 30 and over market is where premium willingness and daytime demand meet. An early evening series with good sound, seating and a strong drinks list is aimed at the guests most willing to pay and least served by late club nights.

The numbers trace a life course: younger people optimise for price and the subway, people in their late twenties stretch the night and miss the train most, and older guests trade late hours for quality. Same city, different constraints, different nights.

Age behaves like the strongest explanatory variable in the dataset, but it is confounded with income, nationality and the researcher’s network. The 40 and over share (29.7%), which is also more expat than Korean (58.8% expat), is the clearest case of sampling shaping the picture. A quota or weighted design in the next wave would separate genuine age effects from who happened to answer.

Atmosphere and music beat price as reasons to choose a venue. For promoters, that means the flyer is not the product; the first ten minutes inside the door are. Bar hopping as the norm also means a night is competing for a slot in a route, not for the whole evening.

Small groups, four to six hours and several stops describe a night built as a sequence. Each move between venues is a decision point where a group can drop off, which is why a bad first room or a long queue costs more than it looks.

The night out here is better described as a route than a destination, which fits time use approaches to leisure. Coding the sequence of venues (start, middle, end, food, home) in a future wave would allow the night to be modelled as a path rather than a set of separate choices.

Food is part of the night, not an afterthought. Partnerships with late kitchens, and a designed non alcoholic menu, both capture spend that currently leaves the venue. The non drinker is a customer waiting for a product.

Ending the night with fried chicken or ramyeon is a ritual, a way of closing the night together. Non drinkers being handed a soda is a small signal that the night was not designed for them, which quietly shortens how long they stay.

The drinks data describes a culture moving from volume drinking toward choice: cocktails above beer, a visible non drinking minority and interest in sober formats. Whether that is a generational shift or a life stage effect cannot be answered with one survey, which makes it a core question for repeat waves.

A modest door charge with a drink included is easier to accept than it looks: almost two thirds of respondents (63.6%) already pay something, and ₩10,000 to 20,000 is the single most common answer. Framed as what the money pays for (the sound, the artist, the room), it becomes a quality signal rather than a barrier.

Free entry lowers the cost of leaving as well as the cost of coming in. When nothing has been paid, walking out after ten minutes costs nothing, so rooms empty quickly and artists play to churn.

Two established ideas help explain this chapter: price as a signal of quality (signalling theory, Spence 1973) and the special pull of a zero price (Shampanier, Mazar and Ariely, 2007). The free entry loop described by respondents is a plausible mechanism, but it comes from qualitative answers and needs venue side data to confirm.

Instagram dominates discovery, yet people still describe finding nights as hard. A well kept, searchable listing of nights by genre and district would solve a problem respondents named directly.

Several patterns are adaptations to constraints: staying out until the first train, moving from the subway to taxis with age, regulars spending less per night. People are working around the system rather than choosing their ideal night.

These cross tabulations are where the dataset is most interesting and most fragile: subgroup sizes shrink quickly, so some differences may not hold. They are best treated as hypotheses, with the next wave designed to test the strongest of them.

Use the personas to plan, not to stereotype: each one is a set of needs a night can be designed around. The Premium Night Out and the Mindful Drinker are the least served and the most likely to respond to something new.

Each persona is held together by a different trade off: music over price, belonging over novelty, comfort over lateness, price over everything, wellbeing over habit. Knowing the trade off tells you what will make that person stay home.

The personas are qualitative syntheses built from cross tabulations and coded comments, not statistical clusters. A future wave with a larger sample could test them with cluster analysis to see which hold up as real segments.