The 20,560 Footprints a Closed Exhibition Left Behind

Saturday, 3 p.m. — the busiest hour in the museum lobby. Someone stands at the kiosk buying a single adult ticket, while at the information desk beside it, a teacher holding children's hands collects a stack of group tickets. That scene looked like a small coincidence on the day. But once we gathered 54 days of it together, it turned out to be no coincidence at all.

The exhibition Yeonmin (SYMPATHY) has closed. Held at the Lee Ungno Museum (a public museum in Daejeon, operated by the Daejeon Goam Art & Culture Foundation) from 2026-04-07 to 2026-06-07, it welcomed visitors over 54 days, Mondays closed included.

Now that the exhibition is over, we turned off the lights and opened up the entire ticketing dataset. Each individual visit was a small choice made on a given day, but gathered across 54 days, they revealed — like a map — how one exhibition actually ran. This piece is that retrospective. Numbers don't make decisions for us, but they do become material for planning the next show.

The exhibition, seen first as a whole

• Total tickets issued (admissions): 20,560, across 13,436 transactions
• About 381 tickets per day on average
• 128 refunds → a refund rate of about 0.95%

Over 54 days, a little more than 20,000 tickets were issued, and the refund rate came in under 1%. Almost no cancellations means that most tickets issued led to actual visits. We looked at these 20,560 tickets along three lines: who issued them, how, and when.

The core insight: the channels divided the work on their own

The clearest thing to emerge was a division of labor between two ticketing channels. The self-service kiosk and the information desk (POS) were absorbing different kinds of visitors, and the boundary between them was drawn sharply in the data.

The kiosk — where individual paid admissions were absorbed

• 11,576 tickets issued = 56.3% of the total
• 1.0 ticket per transaction (one ticket per person)
• Of these, adult tickets numbered 10,396, or 89.8% of kiosk issuance

The kiosk handled almost all of the simple, repetitive issuance — someone arriving alone and buying a single adult ticket. The figure of 1.0 ticket per transaction captures that character exactly. That the great majority of adult tickets were issued here means most of the "come alone, buy a ticket" visiting was processed quietly, in front of a machine.

The information desk (POS) — the place for eligibility checks and groups

• 8,984 tickets issued = 43.7%
• 1,860 transactions, averaging 4.8 tickets each
• Of these, free/concession (zero-won) tickets numbered 6,806, or 75.8% of POS issuance; group tickets numbered 1,249

The information desk, by contrast, averaged 4.8 tickets per transaction — a counter where many tickets changed hands at once. More than three-quarters of its issuance was zero-won tickets. Issuance that requires an eligibility check — seniors, visitors with disabilities, persons of national merit, children aged 8 and under — as well as group visits, are tasks that must pass through human hands, and that work gathered squarely at the desk.

This division shows up in the composition of issuance, too. The kiosk was overwhelmingly single adult tickets (89.8% adult); the desk was mostly free/concession and groups, with zero-won tickets alone at 75.8%. Even the same "one ticket" meant something entirely different at the two counters.

Visiting patterns: concentrated on weekends and afternoons

By day of week

• Weekends (Sat + Sun): 10,353 tickets = 50.4% (weekdays: 10,207)
• Saturday was the highest at 6,036 tickets
• A weekend day was about twice a weekday

Weekends are a minority of the 54 days, yet half of all issuance concentrated into those two days. That the density of a single weekend day is about double a weekday's offers a clue about where to place staff and guidance.

By time of day

• The peak fell in the afternoon, 2–4 p.m. (14:00–16:00)
• Doors open at 10 a.m.; numbers dropped sharply after 5 p.m.

There was a rhythm within the day, too. Doors opened in the morning, the count peaked at 2–4 p.m., then fell quickly after 5. The "Saturday lobby at 3 p.m." we sketched at the start was, statistically, the single most crowded point.

Busiest days

• 2026-05-02 (Sat) was the highest at 937 tickets — early-May holidays and Family Month visits concentrated
• On 5/8, 802 tickets came from 165 transactions (avg. 4.9 per transaction) → likely a day of group visits such as schools

Even "busy days" differed in character. 5/2 was a day of overlapping individual visits; 5/8 was a group-visit day with a high ticket count per transaction. The transaction structure catches a texture that totals alone would miss.

Ticket types, and the place of a public museum

Broken out by ticket count, the types look like this.

• Adult 11,227 (54.6% of the total)
• Senior concession 3,957
• Children & youth 1,278 / Aged 8 and under 1,278
• Children & youth group 1,036
• Other 629
• Concession for visitors with disabilities 491
• Dream Tree Love Card 245
• Adult group 213
• Membership 116
• National merit concession 90

Adult tickets are just over half, and the other half is densely packed with concession, group, and child admissions. Free and concession admissions reach about one-third of the total (roughly 6,600–6,800 free admissions).

That free and concession issuance for seniors, visitors with disabilities, persons of national merit, and children reaches this scale is data proof that this museum is, before it is "a place that charges admission," "a place open to anyone." And as we saw earlier, much of this one-third was issued at the information desk after an eligibility check. The channel division of labor and the role of a public museum turn out to be two sides of the same story.

For reference, this exhibition had the AI docent feature switched on, but as it was in a trial and free stage, we did not separately tally usage figures. So in this retrospective we say nothing about how the docent affected visits. Separating what we know from what we don't is also part of a data retrospective. This time, we focused only on what the ticketing data actually tells us.

What a closed exhibition passes to the next

54 days, 20,560 tickets. Gathering the footprints one exhibition left behind, a few things became clear. The kiosk absorbed the paid and the individual; the desk absorbed eligibility checks and groups; visits concentrated on weekends and afternoons; and one in three people saw the exhibition for free.

These numbers don't make decisions for us. But they do give grounds for questions like how many kiosks to place at the next exhibition, how to staff weekends, and how to open a group-reservation channel. The data a closed exhibition leaves behind becomes planning material for an exhibition not yet opened.

About TixPass

TixPass is an integrated SaaS platform for ticketing, self-service entry management, and visitor analytics — built for cultural venues and MICE, from performances and exhibitions to fairs and museums. It offers dynamic-QR mobile tickets, an AI smart docent, and the Ichnos visitor-flow analytics solution, and is in operation at venues including the Lee Ungno Museum and SETEC.

Contact: sales@tixpass.co.kr