Thanks for the message. I won't go into this and the other aspects of
the current situation in detail -- I think this is an important
conversation primarily with current staff and active community members
--, but I'll respond to a couple points that I think are important,
and for which I can provide some historical perspective.
> In the past year we managed -- for the first time since 2007 -- to finally
> stem the editor decline.
This is a pretty powerful statement! As many folks know, "stemming the
editor decline" was long a top organizational priority, due to
research that showed an increasing tendency for new editors to
encounter barriers, such as the Editor Trends Study, summarized here:
Many will remember the graph illustrating this study, which
specifically underscores that new editors' 1-year retention decreasing
dramatically during Wikipedia's most rapid growth, and remained low
As a consequence, an important number to pay attention to when
characterizing the editor decline is the number of new editors who
successfully join the project. Has that number increased or
It has not, as far as I can tell:
Note in interpreting all data that January is a seasonal recovery
month in editor statistics.
One number to look at here is "New editors", which is the number of
editors who have crossed the threshold of 10 edits in a given month.
For all Wikipedias combined, this number has been in the 12000-13000s
for the last 6 months. Near as I can tell, the last time it has
hovered around or below those levels for this long was a decade ago,
in December 2005. The more modern metric of "new editor activation"
(which does not seem to have the same level of data-completeness)
appears to show similar troubling signs:
Another key metric we paid attention to is the "Active Editors"
number, which has stagnated for a long time; it appears to continue to
do so with no recovery. The most complete visualization I was able to
find is still the one we created years ago, here:
Finally, there's the measure of "very active editors". These are folks
who make 100 edits/month, and one could also call this the "core
community". It's a measure less affected by new user barriers, and
more by the effectiveness of existing editing/curation tools. This is
one metric which does indeed show a positive trend, as was noted here:
This graph focuses on English Wikipedia; this table contains the
numbers for all languages combined, in the "Very active editors"
The numbers for "very active editors" appear to have stabilized at a
slightly higher level than previously. I can't find any firm
conclusion on what has caused this in Wikimedia's public
communications, but the HHVM rollout, long-planned and implemented in
December 2014 under Ori Livneh's leadership seems like a plausible
It seems reasonable to assume that very active editors would most
benefit from performance improvements.
One very positive trend is the Content Translation tool, and its
impact on new article creation, especially in combination with
targeted calls to action, as detailed here:
But overall, it seems premature of speaking of "stemming the decline",
unless I'm missing something (entirely possible). I don't mean to be
negative about it -- I do think it's a super-important problem, and
hence important to be clear and precise about where we are in
> In practice this means I demanded that we set standards for staff
> communication with our community to be professional and respectful. It
> meant transitioning people, shutting down pet projects
Like Brion, I'm also curious what this ("pet projects") refers to.
With regard to tech, I'm not aware of any major projects that were
shut down. I read that major feature development on Flow was
suspended, but active maintenance work to support an active trial
(launched after said announcement) on user talk pages is ongoing, as
far as I can tell:
To be clear, the course of action taken here -- to evaluate a
controversial tool for a specific use case, and see how it fares --
seems completely reasonable to me. I'm just curious if that's what
you're referring to, though, or if there are other examples, perhaps
outside engineering, you have in mind?
Wikimedia-l mailing list, guidelines at:
New messages to: Wikimediaemail@example.com