Hoi,
The biggest problem of this strategy document that I see is that it is
oriented towards our own internals. This is best understood from the first
sentence: *People-centeredness means that every aspect of our Movement must
address the needs and challenges of the people who power it and whom it
serves, so that each one can contribute in their best way to the sum of
human knowledge.*
It does not mention what we do and why we do it. It follows that based on
this we cannot give priority to our biggest bias; what we do first and
foremost is service wants in English. This bias is huge and when we were a
company, we would recognise that around 50% of Wikipedia traffic is in
English. We would realise the extend that scholarly publications are in
English studying aspects of English Wikipedia. We would know that we do not
have much data on everything else. We would be aware of our other products
and strategise how to improve their market value. For our movement, market
value is in the number of people it serves. We would for instance know that
Wikisource books are marketed in India external to us and we would consider
what it takes to provide a proper interface so that people find what is
available to them thanks to a non-English community. A community we do not
notice nor respect.
We are so happy with American students (doing good work) on English
Wikipedia but we do not engage high school students, even primary school
students who could write in their language expanding Wikipedias often with
less than 10.000 articles.
We have an opportunity to turn this around. We have this notion that we are
going to do things differently in this strategy. We have the papers that
for many are too long to read and we have the Special:MediaSearch (publicly
available for two weeks now) that enables search in Commons in all our
languages. When we support it in the Wikiway, we will allow for it not to
be perfect. We will find that as we add items to pictures that we will find
more results or even only results.. Try to find هيلين كوبر using text based
search in Commons and compare the results.
As a product, Commons only serves our own needs. We do not know the number
of downloads of pictures we do not know the extend Commons has been
searched in other languages. This is true for Wikidata as well. We may know
the volume of queries it serves but in what language and how do we extend
the usefulness of Wikidata in languages other than English? What strategies
are in place is this a key performance indicator? How can we show that we
care?
With Commons enabled for search in any language thanks to the
Special:MediaSearch, we have the perfect tool to start address this bias.
We can measure in what language Commons is searched. We can measure the
number of labels added to Wikidata that help people find images. We can
measure the number of downloads from Commons that happen as a result. We
can then demonstrate the pent up need there is.
This will likely be very much driven by the Wikimedia Foundation itself.
There will be outcries from vested interests that it detracts from
other/their priorities. People will state that they are disgusted with us
giving priority in this way. But do realise, white black of yellow, when
your language is English you are well served and others are not. English is
only about 50% of our traffic and you would not say so from what we
advertise we do.
Thanks,
GerardM