Monthly active editors, defined as the number of registered users who made at least five edits in a calendar month, has long been used as a key barometer of the health of the Wikimedia community.
The editor model expands this single metric into a family of metrics that more fully describe who these active editors are and how they became active. The family shares several important parameters:
- Each metric is calculated using calendar months. It's also possible to use rolling 30-day periods, which avoids the problem of different-length months and allows the metrics to be calculated more than once a month. However, these benefits are fairly small (for example, calculating a rolling 30-day metric every day doesn't make it as responsive as a daily metric), and using calendar months makes it more efficient to calculate.
- 5 edits is threshold for being an active editor. This choice is relatively arbitrary, but has been used for a long time.
- All edits are counted, no matter what namespace or page they occurred in. Edits to deleted pages (that is, those in the archive table) are counted as well.
- Bot accounts (as identified via user group membership) are excluded.
The core metric in the editor model is monthly active editors—the number of editors who made at least 5 edits in a calendar month.
Types of active editors
The editor model break active editors down into four subclasses.
|Active last month
|Not active last month
|Registered this month
|Registered last month
|Surviving new editor
|Registered before last month
|Recurring old editor
A new active editors is a user who both registered and made 5 edits within the current month.
Surviving new editors
A surviving new active editor is a user who was a new active editor last month, and has continued to be an active editor this month.
Recurring old editors
A recurring old active editor is a user who registered more than two months ago, was active last month, and has continued to be active this month.
A reactivated editor is a user who was active this month, but was not active last month despite already being registered.
New editor activation rate
The proportion of this month's newly registered users who became new active editors.
New active survival rate
The proportion of the previous month's new active editors who remained active this month (becoming surviving new active editors)
Old active survival rate
The proportion of the previous month's other active editors (that is, those who were not new active editors) who remained active this month (becoming recurring old active editors).