Some of these things are more difficult to test than others, and indeed some are easier to
resolve than others. I'm pretty sure that we lose a lot of new editors due to edit
conflicts. I suspect we can define the people who become active editors as being the
people who learn how to resolve edit conflicts without losing their edit. Unfortunately
there are no public logs of edit conflicts, but it should be possible to create such logs
and test how predictive they are of people stopping editing. If such research produced the
expected result that this was one of the major reasons why we lose editors, then there are
some minor fixes that have been languishing for years in phabricator and its predecessors
so we could easily halve the number of edit conflicts. If the research showed that edit
conflicts weren't driving people away from the pedia then we would have learned
something surprising, and that is always a good thing.
At the other end of the transition scale we have a very very long tail of occasional
editors. I suspect there is a large group of people among them who think of themselves as
Wikipedia users but who will fix the odd typo or other error if they come across it.
I'm assuming such individual editors now edit more rarely as they encounter fewer
typos etc on Wikipedia. Rather than worry that these editors are editing more rarely, I
would like to find a way of measuring such a group that lets us count their finding fewer
typos per hundred hours of reading as a positive sign of quality improvement rather than
as a decline in editing numbers.
Regards
Jonathan
On 23 Mar 2017, at 06:06, Kerry Raymond <kerry.raymond(a)gmail.com> wrote:
A few years
ago the WMF did a survey of former editors, partly to
learn why they'd left. One of the most common responses was "I haven't left
yet".
With the benefit of hindsight (a wonderful thing), that might be a bad way to have asked
the question. A better way might have been to ask why they are no longer active and what
circumstances/change would be likely to make them active again. What we really want to
know if the reasons for inactivity are internal/external to Wikipedia and whether the
conditions for re-engagement are internal/external to Wikipedia. And for the internal
ones, we'd like to know more specifically what they are.
"I haven't left yet, but as soon as my new baby has started school, I might have
the time for Wikipedia again" (i.e. the cause of inactivity and return to activity
is outside of Wikipedia's control). There is not a lot Wikipedia can do about such a
contributors.
"I left because I was sick and tired of the unpleasant way people behave, but I
enjoyed contributing otherwise and would do so again if the culture was a lot nicer"
is something that WP has some control over but not something you can fix in an afternoon.
"I left because I just found it too hard, I kept forgetting when to use [[ and when
to use {{ and I never figured out that <ref> thing" is someone that we could
potentially re-engage on the spot by saying "hey, try the Visual Editor!".
Or maybe "I haven't left yet" is more literally true than we think. It is
possible that the person is still active on Wikipedia but under a different user name or
as an IP so they just appear to have become inactive under their former user name. If a
person has had some unpleasant experiences on Wikipedia and that is why they became
inactive, there are a lot of good reasons why they might not like to return under the same
user name. Wikipedia has an infinitely long memory for things like bans and blocks and
watch lists last forever. If you got yourself in trouble previously but you want to start
afresh, you probably want to create a new account. If you had bad experiences with some
other user who was regularly unpleasant to you, you would want a new account as they can
watch your User page and Talk page forever to detect if you ever return. *Changing* your
user name doesn't solve that problem, creating a new account does. And of course you
may just have forgotten your username or your password and created a new account.
Personally, I am inclined to think that the "I haven't left yet" editors
(who aren't active under another user name) are probably effectively lost to us. Some
other interest has almost certainly chewed up their spare time during their absence from
Wikipedia. There's a big gap between "I'm not saying No" to
"I'm saying Yes".
The other issue is that even if the desired circumstances for re-engagement are in place,
you still need some kind of way to communicate this fact to the "lost users".
Given that providing an email address isn’t mandatory on creating an account, we can only
communicate with those who did provide an email address and hope it is still an active
one.
For example, perhaps we should be emailing all the "lost users" (where we can)
periodically and saying "Hey, try that Visual Editor" or "get involved with
#1Lib1Ref" or mentioning some other positive thing that might convince them to give
it another go.
It's been said (and I really don't know if it's true) that people respond
better to being needed than to being wanted. Maybe we can use that in Project Boomerang.
Find an article that the lost user has made a lot of contributions to but which hasn't
grown much since (ignoring all the re-categorisations, MoS enforcements, reverted
vandalisms, and other edits that don't greatly enhance the information content of an
article) and tell them that article XYZ needs them to come and keep it up-to-date.
In sales, they often say it is 10x the effort to get a new customer than to retain an
existing one. Maybe instead of putting effort into onboarding new users (who we have to
put through a massive learning curve very fast or watch them die the slow death of many
reverts and AfC rejections), we should put more effort into re-engaging lost users
(there's less of a learning curve to bring them back).
Kerry
_______________________________________________
Wiki-research-l mailing list
Wiki-research-l(a)lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wiki-research-l