Kyoto IT Community
Whitepaper 2026

Headline figures

Denominators: the repeat rate and the three-or-more and five-or-more rates are out of 459 unique registrants; the first-time registration share is out of 1,355 valid registrations.

Executive summary

• The frequency of events increased sharply from 2025. We held 28 events in 2019 and 44 in 2025. Activity contracted between 2020 and 2021, recovered from 2022, and has stayed at a high frequency since 2025.
• About 38% registered more than once. Of 459 unique registrants, 174 registered for two or more events. This is based on connpass registrations rather than actual attendance, but it indicates that contact is not limited to a single occasion and that continuing points of contact are being formed. 109 people (23.7%) registered three or more times, and 56 (12.2%) five or more times.
• First-time and repeat registrations coexist. Of 1,355 valid registrations, 459are each person's first registration, giving a first-time share of 33.9%.
• Technical themes and social formats are run separately. Alongside themed events on AI, Web3, web development, mathematics and quantum computing, and reading and discussion, we run continuing co-working study sessions and networking events. Separating theme from format lets specialist learning and cross-disciplinary contact coexist.

Methodology

Primary sources

• Events: mit.connpass.com/event/ (Japanese)
• Members and participation counts: mit.connpass.com/participation/ (Japanese)
• Per-event registrations: the participant page of each event (Japanese)
• Collaborations and outcomes: event descriptions, event reports, outcome pages and official records of the organizations involved

Counting rules

• 162 listed events, 160 held, 2 upcoming
• registered and presenter count as valid registrations
• cancelled and waitlist are excluded
• 529 unknown rows were judged to be extraction noise from organizer and administrator areas, and excluded
• Duplicates of the same anonymized ID within the same event are removed
• No personal names, usernames or profile URLs are published
• connpass registrations are not treated as a count of actual attendees

Figures we do not publish

Events by year

Events by theme

Events by format

The venues we have chosen

We chose our venues to fit the aim of each event, the number of participants, and the balance between focused work and socializing. Beyond temples and machiya townhouses, we have used universities, government and industry support facilities, lodging, coworking spaces, company tech offices, a brewery, the outdoors and online.

To describe our venues accurately, we counted the venue type, the region, the participation mode, and the specific venue name separately. We do not mix the region into the venue type, or blur venues we can confirm into a vague catch-all. We classified these ourselves by re-reading each event's title, description and records. If we revise the classification, we update the dataset version.

8.1 Venue type

8.2 Region

8.3 Participation mode

8.4 Specific venues

8.5 Every event and its venue (162)

Registrations and repeat participation

The repeat rate is not an attendance rate. It is the share of people who registered for two or more events on connpass.

Collaborations to date

This whitepaper does not call every organization a "partner". We record the actual role: co-hosted, co-organized, venue provider, event collaboration, organizer of an external event, and so on. The same organization may have a different role for different events.

Collaborations to date (CSV)

Fuller profiles of each organization are collected on the Collaborations to date page.

Projects and derived activities

We use the phrase "born from the community" only where a concrete causal link can be confirmed. For each category we separate the confirmed outcome from the role Miyako de IT played.

Projects and derived activities (CSV)

What continuous operation has taught us

Limitations of the data

1. Registration on connpass does not show actual attendance on the day.
2. Four events have no participant page, and we do not include them in the registration totals.
3. We have not collected attribute data on a comparable basis across the whole period, so we do not publish the student / working-professional ratio. We also do not infer it from profiles.
4. We assigned a venue name, venue type, region and participation mode to all 162 events, but we may refine the classification of themes, formats and venue types further.
5. 2026 is a partial-year count, as of July 17, 2026.
6. We do not consider that this data alone can prove a causal effect on the regional economy or on careers.

Download the data (CSV and JSON)

The figures on this page and the downloadable files are generated from the same source-of-truth data (version 2026.07.17-v2). Individual registration histories are not published. Column headers in the CSV files are in English; the data values are in Japanese, matching the source data.

• Events and registrations by yearListed events, events held, valid registrations, active unique registrants, new unique registrants, first-time registration share and within-year repeat rate, by yearannual_metrics.csv
• Counts by themeEvent counts and shares by main themetheme_counts.csv
• Counts by formatEvent counts and shares by formatformat_counts.csv
• Counts by venue typeCounts and shares by venue type (the 122 on-site and hybrid events as the denominator)venue_type_counts.csv
• Counts by regionCounts and shares by region (the 122 on-site and hybrid events as the denominator)region_counts.csv
• Counts by participation modeOn-site, online and hybrid counts and shares (the 160 events held as the denominator)participation_mode_counts.csv
• Counts by venue nameEvent counts per specific venue (the 122 on-site and hybrid events as the denominator)venue_name_counts.csv
• Per-event venue data (162 events)Venue name, venue type, region, participation mode and classification basis for all 162 eventsevent_venues.csv
• Collaborations to dateRelationship type, scope and source URL for each organization and venuecollaborations.csv
• Projects and derived activitiesCategory, name, origin, confirmed outcome, Miyako de IT’s role, and source URLoutcomes.csv
• Full dataset (JSON)The source-of-truth JSON containing headline figures, methodology, yearly trend, classifications, collaborations and outcomeswhitepaper-2026.json

How to cite

Netsujo Inc. / Miyako de IT, "Kyoto IT Community Whitepaper 2026", 2026.

If you use this data, please link to this page (https://www.miyakodeit.com/en/whitepaper/2026).

The published data (CSV and JSON) is licensed under Creative Commons Attribution 4.0 International (CC BY 4.0). You may reuse it, including commercially and with modification, as long as you credit the source.

Related pages