We have been organizing Wiki Loves Monuments for three years in a row now, and recently Erik Zachte published some really interesting data regarding how effective Wiki Loves Monuments is in involving more new contributors to Wikimedia projects. In this blog post I will try to dive a little more into those numbers, and paint an image what this looks like in the longer term — especially where it comes to impact on Wikimedia projects. Disclaimer: I’m not a statistician, and the definitions are not always 100% perfect. For this analysis the raw data provided by Erik Zachte has been used — for which I’m very grateful.
Warning: Lots of numbers ahead. Basic take away for me: Wiki Loves Monuments has had a long term real impact on Wikimedia projects and their communities.
First, let’s take a look at the impact Wiki Loves Monuments 2010 has had. In September 2010, the competition was organized in the Netherlands for the very first time. More than 12,500 images were uploaded, by 208 people. 125 (60%) of those users never edited any Wikimedia project before the month of the contest — these would be considered ‘new participants’ for our measurements. Of these 125 people, 27 (22%) were active long after the contest was finished (November 2010 or later).
In September 2011 the contest was organized again, but now at a larger scale: 18 European countries participated. This contest was the largest photography competition organized until then, with 168,208 submissions by 5,368 people. The percentage of ‘new participants’ was even higher: 65% (3,497). 11% of these remained active, making edits in November 2011 or later.
In September 2012 the contest was organized for the third time — now in 35 countries all over the world. The numbers of the previous editions were quickly forgotten as the 2012 competition received more than double the amount: 350,000+ images were submitted by 15,110 people. Again we counted how many of those people made their first edit during September 2012: 10,825 (71%) had never edited before. The percentage was especially high in some countries, such as India (no exact numbers at this moment). Of these new participants, 618 (5.7%) edited in November 2012 or later.
It would of course be nice if we were able to compare the different contests, and see how the results might scale. Because it will be obvious that the chances that someone has edited again increase over time. We tried to compare them, by only looking at the new participants that edited in the last six months (December 2012 – May 2013). To our very surprise these numbers were quite consistent. If we look back at the 2010 contest, 125 new editors joined during Wiki Loves Monuments. 25 (20%) have edited since November after the contest, and 5 of these users made at least one edit in the past 6 months – 4% of the new users who joined during Wiki Loves Monuments 2010 have returned in 2013. If we use the same definition for 2011, we find out that 11.3% edited since November after the contest, and 140 people (again 4%) have edited in the past 6 months. Now if we look at the 2012 competition, much more recent, we find that of the 10,825 new participants, 618 (5.7%) have edited in the past 6 months.
Also very interesting is to take some cumulative numbers. With Wiki Loves Monuments, measured over the past three years, 14,447 people who had never edited before uploaded an image, and 1,042 went on to edit one of the Wikimedia projects after the contest was finished! And even more promising: 626 have remained active, and made an edit in the past 6 months. This might be a relatively small number of users compared to the whole Wikimedia community, but it is still significant. But maybe even the most exciting and inspiring number: all these new users together have contributed 616,138 edits to the Wikimedia projects, including their image uploads. As we see that about 4% of the new participants continues to edit over a six month period, this number will probably only increase further.
I find these very promising and surprisingly positive numbers. We already knew that large amounts of content were collected with this type of contest, and we also knew already that this is one very effective way of reaching out to large amounts of new contributors. We saw a big spike in the number of uploads and the number of new accounts registered during the competition months on Wikimedia Commons. What we didn’t know for sure until now, was that this is a lasting effect, that there is a long term impact on the communities. And what is the most encouranging to me personally is the 4% that made edits in the past six months, independent of the year they joined.
NOTES: A few inaccuracies are notable:
* For 2011, Switzerland actually started already uploading in July. New participants who joined in July/August 2011, are here counted as ‘old uploaders’.
* Another inaccuracy exists for Hungary in 2011 and Israel in 2012, which started halfway September and continued into October.
* People who have only uploaded an image to Wiki Loves Monuments that has been deleted, are not taken into account – no matter if they edited afterwards.
* In these numbers, an upload is considered an edit. Edits to any project are added up for a given username.
* For future investigation (feel free to add more suggestions in the comments): exactly how many edits were made by these new participants *not counting* the contest. How many of these edits by ‘WLM-rentention users’ were made on which projects? How many participants from 2011, 2012 were ‘new participants’ in the years before? How many people
* I’d love to make this more insightful with helpful graphs, might add those later.