Cultural Conversations
Cultural Conversations

What citizens are asking their cities about culture

City-deployed AI chatbots, fed by Encore. Each conversation is a real citizen asking a real question about their city's cultural programme — we aggregate, classify, and report.

Monthly insights · May 2026

May — the cohort steps up.

137 cultural conversations across the Capital of Culture network. 78 distinct sessions across 25 active days of 31. Five cities, five languages handled natively — English, Spanish, Czech, Dutch, Greek. Average response length 1,341 characters; 85% of replies are substantive long answers.

Messages023%vs Apr
Sessions013%vs Apr
Cities025%vs Apr
Languages0+17%vs Apr

Five cities live: Brno, Cáceres 2031, Larnaka 2030, LOV2030, Novi Sad. Cohort mean Q/session 1.76 (a June reading above 2.0 would suggest a move from probe to exploration). 35% of queries are ≤3 words — keyword-recognition strength is load-bearing for the whole cohort.

City by citySorted by share of conversations
  • LOV2030Belgium

    Dominates volume (52% of cohort messages) at average depth — 1.65 questions per session, broadly matching Brno. A broader audience asking more questions, mostly shallowly.

    52%
  • Cáceres 2031Spain

    Punches above its weight on engagement depth. Only 29 messages but the deepest sessions of the cohort: 3.2 Q/session, both of May's 6Q+ deep dives, and the longest answers on average.

    21%
  • BrnoCzech Republic

    Steady mid-volume traffic; response length (770 chars avg) the cohort outlier — substantially shorter than LOV2030 (1,301) or Cáceres (2,024). Worth checking whether the source documents are leaner or whether the agent is tuned more concisely on purpose.

    15%
  • LarnakaCyprus

    Highest one-shot rate of the cohort — 58% of queries are ≤3 words ('synikismos', 'open call', 'Dance'). Keyword-recognition strength matters most here.

    9%
  • Novi SadSerbia

    Smallest volume but uniformly substantive — every reply categorised long-form (100%), 1,289 chars average.

    3%
Per-city scorecardDepth + response metrics
CityMsgsSessDaysAvg QMax QDeep (6Q+)Avg resp (ch)Long resp
LOV20307143191.655—1,30185%
Cáceres 203129963.201222,02497%
Brno211371.604—77067%
Larnaka121191.092—94292%
Novi Sad4222.002—1,289100%
Cohort total13778251.751221,34186%
Session depth

How many questions a session contained.

  • 1Q47 sessions (60%)
  • 2Q20 sessions (26%)
  • 3Q4 sessions (5%)
  • 4Q5 sessions (6%)
  • 5Q0 sessions (0%)
  • 6Q+2 sessions (3%)

86% of cohort sessions are 1–2 questions. The cohort is largely a probe-and-go audience.

Daily activity

25 of 31 days had cohort activity.

  • Peak day: 22 May — 21 messages, Cáceres + Larnaka concentration.
  • Secondary peaks: 18 May (13 messages) on the LOV2030 Summit eve; 28 May (13 messages) LOV2030 post-Summit + Larnaka.
  • Quietest stretch: 11–14 May, ~1 message per day.
Language mixBy city, this month
  • Brno

    Predominantly Czech, with English on policy queries.

  • Cáceres 2031

    ~80% Spanish, ~20% English.

  • Larnaka

    ~75% English, ~25% Greek (incl. transliterations). No Polish this month (23% in April).

  • LOV2030

    ~65% English, ~31% Dutch, with one Czech surprise — possibly a cross-city user testing the widget.

  • Novi Sad

    100% English.

What people asked aboutTopic distribution this month
  • 10%
    Strategy & bidbook13 hits (9.5%). The cohort's clearest engagement spine — the content these agents are best equipped to answer well.
  • 8%
    Events & dates11 hits (8.0%). Time-critical event content drove the network's biggest engagement days.
  • 6%
    Brand assets & logo8 hits (5.8%). Logo, brand-use, and downloadable-asset queries cluster across LOV2030 but appear across the cohort — a candidate for a shared brand-asset hub pattern.
  • 5%
    Contact & staff7 hits (5.1%). Who-do-I-talk-to is a reliable bottom-of-funnel signal.
  • 4%
    Vacancies & jobs6 hits (4.4%). Cluster in LOV2030 and Larnaka — the chatbot is quietly becoming a first point of contact for employment.
  • 4%
    Open calls & submissions6 hits (4.4%). Practitioners (artists, partners, prospective volunteers) showing up to ask how to participate.
  • 2%
    Volunteer & participation3 hits (2.2%). The 'how do I get involved' content surface is the next frontier.
  • 2%
    Partnership3 hits (2.2%). Cross-city and inter-org partnership questions surface modestly.
Evidence in the making

The case for an 18th SDG on Culture

Look at what people are actually asking these chatbots about — and you start to see something interesting. Mapped against the 17 UN Sustainable Development Goals, the conversations don't scatter randomly. They cluster. Culture isn't a sidebar to sustainable development. It's the thread running through almost all of it.

  • SDG 11

    Sustainable Cities & Communities

    Infrastructure, cultural-centre plans and accessibility-across-the-whole-city questions continue to lead.

  • SDG 17

    Partnerships for the Goals

    Cross-city twinning, European co-productions, partnership and inter-municipal questions persist across all five cities.

  • SDG 4

    Quality Education

    Cultural-education programmes, youth cultural work, and capacity-building enquiries.

  • SDG 10

    Reduced Inequalities

    Multilingual access still uniformly strong — five languages, zero fall-throughs reported.

  • SDG 16

    Peace, Justice & Strong Institutions

    Intercultural dialogue and trust-building themes — particularly visible in the deeper Cáceres sessions.

  • SDG 13

    Climate Action

    Climate-and-culture overlap remains a steady minor share — concentrated in LOV2030's sustainability-themed exchanges.

The missing goal

SDG 18: Culture as a Driver of Sustainable Development

Every conversation in this dataset touches at least one SDG. Many touch four or five. The people asking these questions aren't thinking in silos — they're asking about belonging, about identity, about what their city stands for and how they can be part of it. Culture isn't the means to the other 17 goals. For many of these users, it isthe goal. The evidence for a dedicated SDG on culture isn't in a policy paper. It's in 489 conversations — and counting.

Things worth noting
  • Working well

    Depth is uneven — and that's a signal

    Strategy & vision content drove the largest single share of conversations across the network. From single-word keyword probes to 12-question deep dives, audiences across the cohort are using these agents very differently — and getting substantive answers either way. 86% of cohort sessions are 1–2 questions; the cohort is largely a probe-and-go audience, but a small slice of deep-dive users (3%, 6Q+) carries a disproportionate share of the qualitative signal.

  • Growing trend

    Practitioners are showing up — quietly

    Artists, partners, prospective volunteers — they're showing up in the data, asking how to participate. The 'how do I get involved' content surface is the next frontier. When source documents on participation are strong, the answers are specific and useful; where they're weak, the gap is visible.

  • The opportunity

    Time-critical content compounds

    Peak day was 22 May (21 messages) — Cáceres + Larnaka concentration. Secondary peaks on 18 May (Summit eve) and 28 May (post-Summit + Larnaka). A consistent pattern: getting fresh content live, on time, pays off — confirmed network-wide. Quietest stretch was 11–14 May (1 message a day), which broadly tracks European bank holidays.

Aggregated from 5 live city widgets · feeds the next Tuesday digest