# What the public Propel event archive records

> Propel's public archive lists 141 rooms from 2025-06-08 to 2026-10-22: 87 in person and 54 online. They record 7,129 registrations in total. The median room recorded 25. These are registrations, not attendance and not unique people.

Canonical: https://propel.earth/
Last updated: 2026-11-05

Propel's public archive lists 141 rooms from 2025-06-08 to 2026-10-22: 87 in person and 54 online. They record 7,129 registrations in total. The median room recorded 25. These are registrations, not attendance and not unique people.

First-party observational report. Snapshot of /api/event-archive taken 2026-11-05. Reviewed 2026-11-05.

Propel Earth runs these events and has a commercial interest in how they look. This is a descriptive report on our own public calendar; it does not show that Propel causes any outcome.

## What it found

- 141 rooms in the archive; the sum of recorded registrations is 7129. The median across all rooms is 25.
- 18 rooms are recorded as 0 registrations. We do not know whether 0 means none or not recorded, so we show the median both ways: 32 among the 123 rooms with a positive count.
- In-person rooms have a median of 21 (24.5 among 74 rooms with a positive count). Online rooms have a median of 33.5 (34 among 49).
- The format difference is descriptive. It does not show that online rooms cause more registrations.

Group · Rooms · Registrations (sum) · Median, all rooms · Median, positive counts only

- All rooms · 141 · 7129 · 25 · 123 rooms: 32
- In person · 87 · 5077 · 21 · 74 rooms: 24.5
- Online · 54 · 2052 · 33.5 · 49 rooms: 34

## Method and source

- Source: the public feed at https://propel.earth/api/event-archive, fetched 2026-11-05. Its underlying calendar is https://luma.com/propelearth. We froze a projection of four fields: start date, in-person or online, Propel's room type and the recorded guest count. No names, hosts, links or guest lists are kept.
- Medians are computed over the recorded counts. With an even number of rooms the median is the mean of the two middle values. The calculation is in shared/data/researchArchive.js and is covered by an offline test.
- This is a first-party observational report. We run these events, so we have an interest in how they look.

## Limitations

- A registration is not attendance, and one person can register for many rooms. The total is not an audience size.
- The archive is Propel's own calendar. It is not every founder event, may not hold every past Propel gathering, and is not a representative sample.
- Room types are Propel's own labels, not an industry taxonomy.
- Calendar coverage in parts of 2025 is unverified, so the dates do not show growth.
- The archive is a live feed. The figures here are a dated snapshot and will drift.

## Questions

### How many people have come to Propel events?

This archive cannot say. It counts registrations, which are neither attendance nor unique people.

### Do online events draw more people?

In this snapshot online rooms have a higher median registration count. That is a description of this calendar, not a format effect.

### Why show the median two ways?

Eighteen rooms are recorded as 0 and the data does not say whether that is a true zero.

## Sources

- [Propel Earth event archive feed](https://propel.earth/api/event-archive)
- [Propel Earth events](https://propel.earth/events)
- [Propel Earth on Luma](https://luma.com/propelearth)

## Related

- [All research](https://propel.earth/research)
- [Events](https://propel.earth/events)