 Yes, okay. Thank you so much Yes, I will be talking a little bit about Wiki Loves a Few different week loves competitions we could last monuments week loves earth and week loves Africa and especially what we can learn from them In regards of numbers, I'll be throwing some statistics at you, but I'll try to focus on the bigger picture And what we can learn from them First of all, I was just curious like who is in the room here So who has organized a Wiki Loves competition in their country? That is actually already like a lot of people that's really good to know and who has ever participated in one of them has submitted a photo To any of the competitions Wow, that is lovely to see So first of all all the images that you see are winners of a Wiki Loves competition international So do look them up if you're interested if you're curious the text is sometimes blocking the pictures. I'm really sorry for that But you'll see definitely a lot of pretty pictures and I'll throw in the link at the very end I'm just to give a quick overview of Of what the today's session is going to be about First I'll quickly recap what Wiki Loves is all about how it started very big picture because you've already heard some from Alicia Then We'll talk a little bit about like the images we collected how they're being used And the people that joined through these competitions into our community So first let's start with a little bit of context In 2009, which was before Wiki Loves monuments. We had Wiki Loves art. It started in the United States In New York and they they started that is why we have Wiki Loves because it was in February and that was like a nice connection Wiki Loves Valentine's Day And they were going to museums and they were trying to take as many pictures of objects in the museum as possible Then in the Netherlands we got feedback from the museums that maybe Wikipedia is running around in the museum to take as many pictures as possible is not the most optimal experience for the museums So we switched to a more qualitative aspect where we had which is where the jury started to come in So in 2009 we did that in the Netherlands with 14 museums and as you can maybe imagine that's a lot of work Signing 40 contracts 40 partnerships that really required a lot of effort and that just did not skill We did not want to do that again So in 2010 we were looking for something a little bit more in the public space where we could organize something Where we did not have to make partnerships And that is how Wiki Loves monuments came to be in 2011 that expanded to Europe and In 2012 Worldwide and before you know we had a Guinness World Record for the largest photo competition in the world In 2013 we saw the first spin-off in Ukraine that you heard a lot about Wiki Loves Earth and In 2014, I believe Wiki Loves Africa started and more and more photo competitions that I'm probably leaving out here But for simplicity's sake I'm going to focus on these three photo competitions which are the biggest of that set With spending the most different countries Wiki Loves Monuments has now been organized 13 times and as you can see many of the countries in the world The darker the shade the more times they have participated Have participated at some point in the competition And you see that it's especially in Europe there has been a lot of participations over the years But a lot of countries in the world have done it to some extent And if you look at Wiki Loves Earth I asked the creator of the previous picture to also create one for Wiki Loves Earth So this is a new one and you see there you see a somewhat similar pattern But definitely different parts of the world being active which is really nice to see because it shows that the organizers of the competition Like they have different focuses and different cultures work better with different topics So the concept of photo competition is pretty similar But the way it's engaged is very different I won't show a map for Wiki Loves Africa because we all know where Africa is and it spends the whole continent The number of countries that participates every year has changed quite a bit So green that is Wiki Loves Monuments, orange is Wiki Loves Earth And you see that the number for Wiki Loves Earth is definitely still going up last year this year I think it has reached peak height Wiki Loves Monuments is going down a little bit. It's more stabilizing at this point Around somewhere between 35 and 50 countries is what we see is fairly common for this type of competitions Which is actually that is a lot that means that you need like 30 at least for Wiki Loves Earth and Wiki Loves Monuments You see 35 teams of that are organizing something in their countries. That's a lot of coordination That's a lot of collaboration which is quite exciting all by itself As you can see we're not only talking about a Like we're talking about a lot of different competitions and every team has like their own goals and their own Their own reasons of why they're organizing it And this is some of the reasons that we most often hear from in in surveys, but also in in individual conversations Of course the images is very important, right? Like getting a lot of images from our heritage around the world in all its different shapes and types Is really valuable because we can illustrate things in our encyclopedia and wiki data in different projects And we can show people what it's all about But it's not just the the quantity that matters, but it's also the quality We want to get like really good pictures for complicated topics and we can use it then in also in different contexts But besides the images itself my personal driving motivation And it's all about the contributors the people that we get on board in our projects through the competition There's these people that that never would have considered editing Wikipedia. Maybe Just because they come on the website They see this banner and they think like hey, that's something I can do I can take a picture of that building that's around the corner I can take this picture from my vacation that I upload that I had taken earlier and Upload it to a photo competition and some of these people stick around so that is my my personal motivation But every team has their own motivations Another topic that another reason why people organize it is sometimes because it's it's helps you to do capacity building It might be you're a you're a new user group. You're a new team somewhere in the world You're trying to find out like what is an activity that engages our audience that engages our country And how can we organize something that is? relatively well structured and wiki loves might be a good place to start if that is Something that interests your audience if you have photographers if you have organizers that might be a good place to start and Finally, there's improvement of heritage data We have so much information in our in our projects, but also wiki data for example We really want to understand like what what monuments do we even have in our country? What natural heritage do we have? And just collecting structuring that information might be valuable all by itself So every country every national team has their own goals of why they are organizing this And I think that is good to just recognize because everybody every team is different. I'm into this. That's amazing So to look at this from a from the perspective of a participant Let's let's take us quickly through a journey of like how someone ends up in wiki loves In most of the cases we see that people they see a banner on top of the wiki page of the Wikipedia, right? They come to the encyclopedia They look something up that is totally unrelated and they see this banner and they say like submit a photo Participate in a wiki loves monuments wiki loves earth and And then they click on that banner the next thing they come to a landing page that explains what the competition is about what they're Supposed to photograph and what they can what they can win perhaps. What is it? What do you have to do? Then the third thing is they need to well I skip a step they first need to figure out what the photograph that's sometimes a challenge depending on the country, but That's definitely a step and then they take the photo they upload the photo They create an account in the process and then sometimes they come back after that Maybe for the next competition maybe to just contribute to Wikipedia because they now realize like hey I can actually do something that makes this website that I use every single day better and and they They they may come back and then we see some kind of retention Maybe to participate in a different kind of competition because they kind of get that vibe of ooh That is kind of fun making a photo where I know that people actually want to have that photo That is quite a satisfying feeling So let's talk about that first phase that we have in that process the banners So thanks to the research team I was able to to look a little bit deeper into this in 2021 and we did a project where we were looking at Trying to figure out like how do people arrive at these landing pages? How many people do that and How do they then get to the uploading phase? So so all these different steps like what happens with these people. There's a lot of caveats in there There's a research report on meta. I'm I'm welcoming you to to look into that if you're interested in that But there's a few things that we definitely learned that might be worth while mentioning So first of all we see about 0.2 to 0.8 percent of the viewers of the people that visit Wikipedia They actually click on the banner and less than 1% of the people who actually click on the banner typically goes then to actually Upload something which is what we consider a success That is there's actually a big range in there. So there's for every country There's a different number of people that actually come from the landing page to an upload and that can range all the Way from 0.08 percent to 1.7 percent So it's really interesting which tells me that the landing page design really really matters that may be obvious sometimes But it's really important that if you if you organize a competition like this You have to talk with some people who are not weak comedians and just say like hey Is this a sensible design? How can we improve this? How can we make sure that people know what to do if they arrive at this landing page? We what we also learn and again, these are some things that are maybe obvious to to some of us and not obvious to others If someone actually visits the landing page They're more than 25 times more likely to actually visit also the account creation page So that means that if someone clicks on that banner, they're also much more likely to then also create an account Compared to people who are just visiting Wikipedia itself If if a reader then visits the landing page, they're also much more likely to also visit the upload page Which is good, right? Like if that would be equally likely then why are we doing even doing the campaign? And finally if someone actually Creates an account through the campaign. They're also much more likely to upload an image than if you create an account Not going through that banner These are just some interesting tidbits. They're not going to change your life But I think they're really telling us that the campaigns that we are running are actually working to get people on board of our projects Another thing that we looked at carefully is like how many banners do we actually need? Like we talk a lot about diets and things like that There's a lot more work to be done here to really understand what's going on But something that we did understand is that the number of people that see more than three banners before they arrived for the first time on the landing page That is about That's about half the visitors so half the readers That do arrive on the landing page have seen more than three banners in that same day Which tells me that three a diet of three banners in a week is maybe a little bit on the low side Because that means that if you would successfully implement the diet you lose like half your visitors Secondly The United States introduced a diet in the middle of the campaign Which we were actually not aware of until after the fact and that's capable some kind of an experiment Unintentionally we saw a significant drop in the number of visitors, which is not surprising given what I just told you But we also saw that there's still quite a lot of visitors that actually still had more than 10 Banner impressions before they arrived in the landing page. So the diet is not working As we would expect it to be and Finally, there's it is very unclean experiment So you would have to do a lot of extra research if we want to do that properly So what did we learn every step of the pipeline? We lose a lot of people. So try to reduce the number of steps Lending the part design of landing page design really matters a lot We could as monuments draws in people in higher rates than through natural recruitment if you just come to the website And showing banners multiple times really does make a difference It might be a very different behavior than what we see in fundraising for example Because in fundraising if you just click on a banner you can immediately take action If you click on a weekly loss monuments banner You first have to actually find a photo before you can upload it or find or take a photo So it may takes a little bit more effort and introduce introducing a diet Is probably not helpful for recruitment purposes, but there might of course be different reasons why you do that So now the question is if people actually are recruited this way like what what happens with them? How many people stick around? I'm sorry I would like to talk you a little bit with these people these these uploaders that never edited before So first of all, it might be good to just get a little bit of context of like how many uploaders are we talking about? So this is the number of uploaders for each competition per year that we see through all the countries combined Of course, if you crane that decides not to participate one year that means a big drop in the number of participants Right, so it really depends on like which countries are participating But this is just to give you a little bit of context of like how many people are we talking about? We do see a significant drop after 2020 So something is changing in our ecosystem and that might be for example that would just have more competing banners Or it might be something else in our infrastructure. We don't really know but it's just interesting as an observation If we look at the number of countries, these are the countries that have generated the most up the most Uploaders over the past years So that's just interesting to see that we really see at India Germany Russia That's the kind of countries that generate the most uploaders again not surprising. They also have the most people living there But what is really interesting is also the number of images per uploader So in we see that in week it was monuments There is a very different kind of behavior the number of images on average per uploader is much higher that may have something to do with Super uploaders. There are some people that upload more than 10,000 pictures per year. It's like, I don't know how they're doing it It's amazing But it definitely skews the numbers a little bit So more research is needed in this if you're wondering why there's always this gap in 2018 for week it loves Africa by the way That's because they shifted months from the end of the year to the start of the year and that is why in 20 So they went from the end of 2016 to the start of 2018 17 and 19. Sorry. Yes So that is and there's something we would expect The other thing that is interesting is just a cumulative fraction of uploaders So if you look at the number of uploads and how many people are part of that You see that there is more people in week it was monuments that upload more images So there's a lot more people in week loves earth and week loves Africa that upload a single or two or three images It's just a different kind of competition. There's nothing wrong with that But that's just as a little bit of context of what kind of competition we're talking about So what do the participants upload And how many and how many of them stick around for that? I looked at all the images in the images from wiki love something category Split it up by country split it out by year and look that the people who are in that set And I looked at their first edit if that edit that first edit that they ever made on any wiki media project Was in the same month as the competition. I consider them a new editor I know it's not a super clean definition But it's something to work with that is what we call month zero and then everything after that is month one month two Etc so all the months that you see later in the presentation They're all relative to their first month of participation and their first edit Now the first thing that that might be interesting to look at is like how many are actually new editors There's a little bit of a bug there week loves Africa does not have one one hundred and six percent new editors in 2019 I know what the buck is coming from but Just as a background But we see that actually there is about 60 70 80 percent new editors consistently over the years for every single competition And then you wonder like ah, but there must be some countries that have a much lower number of new editors, right? What turns out that's not the case. This is every single country activity combination and you see that actually Every single time it's above 60 percent new editors And that's really exciting to see that if you use this banner if you use this this format You actually get a lot of new editors every time So then the question is like what what do we see if we look at like how many of these people stick around? So we look at like how many edits they made in the months after and if they made at least one edit I consider them to be active. So it's a very low but a threshold But it gives you a little bit cleaner numbers and if you use higher Thresholds if you use a higher threshold of five at us, for example, you get similar patterns But it's just a little bit a little bit more noisy and what you see is 2010 was a very different here than all the other years Which makes sense. It was just in the Netherlands and the first year that you organize one of these things It looks like the first year is always a little bit different because you're you're basically taking the low-hanging fruit You're catching all the people that are super excited about this and that really get engaged And they stick around much more than the people later in later years But after that you see that it kind of stabilizes the number of people that make at least one edit every month Is actually quite reasonable. It's somewhere around 1% What we also see is there is a big peak 12 months later That is not surprising for people who have organized this competition probably But that is basically the people who return to the next year's competition So there's actually quite a number of people that return to the next year Which is two three percent depending on the year depending on the country But that's really interesting a peak that we maybe expected a little less is a peak roughly about eight nine months later And that has to do with week loves earth So this is the week-loss monuments data and we see week-loss earth is about eight nine months later And you see a little peak there which is the people that basically cross they joined through week-loss monuments And then the next year they joined week-loss earth So that's really interesting to see and you see that those peaks go down over the years But people keep coming back a year after year But there's also between the competitions. They're still quite a bit of retention We see roughly the same patterns in other years the numbers are a bit lower the percentages are a little bit lower But you see roughly the same peaks the same kind of movements If we look at week loves Africa or week loves earth we see similar patterns, of course the eight-month peak is now a three-month peak But you see roughly the same patterns We could love Africa is much more noisy And it probably has something to do with the fact that we close Africa is a bit more spread out over the months It's a little bit more every country is different and maybe does not have a week-loss earth or week-loss monuments competition So the peaks are a little bit more noisy But in both cases you see that there's actually still a fair number of retention, but it's different patterns So if you really want to understand that we have to dig deeper. I didn't have time to do that But I would really invite the organizers of those competitions to do that together What we see is if we look at multiple months, so we take a ten month period So we take for example if you have a competition in September That's month one a month zero then month one is October and then we look at November all the way through August That's a ten month period and we look at like how many people make at least one edit in one of those months And then we see that there's about five point four percent in the case of week-loss monuments That make at least one edit in that ten month period. So that means that these people at least have their passwords still That's a very encouraging thought for me because they can be reactivated Because you get logged out about after I think after one month you get logged out. Is that still the case? Someone here must know that Sorry oh one year is now it is okay So it could it so we have to look at after one year. So I'll try to rerun that at some point But basically five point four percent of if you look at their whole lifetime So months to until eternity so that includes the next edition that we see it's about ten percent 11% for week-loss monuments week loves Africa has about 10% and for we could observe it is about 7% Of course these numbers are highly affected by the fact like how hard do you make it to join in the first place? If you make it very easy to join that percent is going to be lower If you make it very hard to join like you have to do a lot of effort to join for it at first time Your retention percentage is going to be through the roof What is also interesting to see is like how many contributions did all these retained users? So we're talking in the case of week-loss monuments 74,000 new editors joined over the course of these 13 years that Like together they made 5.7 million edits after they joined so in months to and later In the case of week loves earth that is 1.7 million edits or contributions in the case of week loves Africa We're still also talking about 138,000 edits. That's a lot of edits For for that number of users and it's really encouraging because it's not just about the competition itself After that you have a lot of effect going on So we also looked at like different countries and what what I find most interesting I mean you could look at these graphs forever, right? But what I find most interesting is there's some countries that have a much higher Retentuated others and what I would like to do with those organizers dive deeper into this like why it's Germany is such a high retention rate What are they doing differently? Are they making it really hard to join or are they making it really easy to stay? And I would like to understand it a little bit better So I think we need to talk with countries that where we see that kind of patterns in every competition You see some differences. So for example, if we look at week loves earth, we see that for example Germany and Germany and Ukraine are having a particularly high retention rate compared to the other countries And I think we need to start learning from each other a little bit more in that context in week loves, Africa I was mostly surprised by Nigeria that was doing particularly. Well, you see that there's a lot of noise But there's one country that has a consistently retention rate And that is Nigeria So that's really exciting that may have something to do with the fact that they have more Activities in that country it may have to do something that they have a very supportive user group We don't know we should dig into that So some observations from all of that from a non-competition month We see roughly 1% retention in the case of week loves monuments week loves Africa She's also about 1% percent a week loves earth see about sees about 0.6% If we look at that 10 month span, we see 5% for week loves monuments 6% for Africa and 3.5% for a week loves earth And if we look at lifetime retention those percentages are even a bit higher There's always a retention bump around 12 months, which is the returning visitors And we see a bump when other competitions are happening So really try to encourage your participants to participate in the other competitions in your country Because it helps them to re-engage Most of the retention happens on Commons I didn't show you the figures for that but about 50 to 60% of the off cycle retention Happens on Commons, but that also means there is there's actually quite a bit of retention happening Outside of Commons. So if we want to look at retention numbers, we have to look in the full breadth of our movement And there's a few smaller countries that have an even higher off cycle retention rate in Germany So I wouldn't know we should talk with Austria. We should talk with the Netherlands to see what what they have done differently over the years So finally it's worthwhile considering contribution across all projects if you want to look into these numbers There's big differences between countries and how can we leverage this reactivation In a better way. So thank you very much for your Participation there's a lot of work to be done in this. There's a lot of more data to dig into and if anyone is interested Yeah, hit me afterwards. I think we have a few minutes for questions, which I would very much invite. Thank you Sure, please Yeah, please quickly introduce like your back like your your name and Yeah, hello, I'm Matty username is MB1 I'm mostly active on comments. So it's very relevant for me about the retention You said the retention as high as on Commons, which is hardly surprising to anyone, I guess But I would be interested. Did you look at the figures of? Uploads versus other types of edits on Commons as well I look I look at the edit table, but every time you make an upload that also generates an edit basically So I would also see those I see them as one package Yes, but the question would be is is editing which is not uploading on Commons also Significantly higher than on on other wikis in the retention. I don't know. That's a good question I have not looked at that separately. Okay. Thank you Yes, if you have questions, please just go to one of the microphones and Hello, I'm Robin from the Philippines and I organize different week allows a contest So I just want to share the reason why there are very low retention of the Philippines is that Once you create an account Using your mobile phone phone, you'll automatically get blocked So that's the main reason why there's a low participation in at least in the Philippines and another one is that the We can call musty when you upload that it's not very User-friendly there are lots of fields to fill in so it discourages participant Thank you. What I would like to say in response is I don't think 1% is low I think it's actually really good. So if we hit 1% retention consistently over this kind of number of new editors We're doing actually really well. So so don't feel like you're doing poorly in that sense But secondly, these are great thoughts of how we can improve our retention experience. So thank you for sharing those Hi, so there's a question from Ina Lee Hi retention in NL. When was the last time they ran a wiki love sex? I don't know. I think it's been a few years and It may have been a while. I Think it may have been three or four years. I could look into the data Use a poco poco on commons You mentioned that one your of your main targets is to get quality content. Have you measured? Quality on the uploads. Is there any metrics about this? I'm some beta or words or so what I wanted to just to clarify These these goals that I listed were goals from local teams. So every local team has their own goals I have not personally measured the quality That's a great thing to look at but I have not looked into that one comment So I'm George jiku from from Moldova and they're organized weekly observed and moments and We saw in the graph that In the graph with the contributors that the graph was Starting to get the number of contributors was lower and lower and lower In our case it was because of purely because of saturation. We we just don't have much to to take photos off anymore because we already filled up the tables so to speak and it requires a lot more effort at this point to To make the the photographers go out in the field again, thank you makes a lot of sense I Think we have one time for one more question and then we have to close up Mike from New Zealand We haven't yet done photographed everything in New Zealand yet. So there's still some time It wouldn't be possible to estimate the person ours per retained Editor based on the calculating the amount of effort involved in administration and judging comp photos divided by the 1% retention and compare that with personnel is expended to retain editors and other things like editor thons So the data that I found is is publicly available here under on the pause Sure, I know that but is there is do the projects collect the hours of Personals in the administration time. So I don't have that data. So what I can do I can offer you one piece of that puzzle I can offer you how many new editors there are if you then want to know how many Hours you need to retain one of these editors then I would suggest that someone else can maybe provide you those hours And you can match that data together Okay, if there is probably more questions. We have a Session later there maybe it might be some time over there But for now, I would like to give the floor to the next speaker. Thank you very much