 Wonderful. Welcome. It is June. It's our June metrics meeting. I'm going to start with the welcomes and the hires. Guess what? We hired a Lila, which is great. And we have a whole bunch of other people who are new to the organization. So as I rattle you off, raise your hand if you're here so we can eyeball who you are and people can know to come over and say hello. So we have Ramarad Sitzman in engineering out of Colorado. Ramarad Daniel Duval. Where is he? There's Daniel. Filippo Giancetti in Ireland. Ramana Yazmin in engineering. Ramana, raise your hand. He's back there. Abby Ripstra in product. We have Makunda Modell in engineering and he is in the office today. Rachel DeCerbo in product. Who is Rachel? Okay. And Elena Hernandez in HR. That's Elena. And then we have a number of new office IT interns, which is fantastic. I'm going to name them all off. Sylvia Ventura, where are you? Not office IT. No? I'm raising. I'm sorry. Ralph Torres. Where's Ralph? Not here yet? And Consuelo Jimenez. Connie. That's Connie. Fantastic. And then we have a bunch of legal interns. Mark Vastrati. I might be butchering these names. Chuck Rosloff, Joseph Young, and Eric Holm. All right. Stand up. All right. Great. I'm going to turn the mic over to Lila, who will start off with an executive director update. Okay. Excellent. Thank you so much. So I promised all of you a while back that I will keep you appraised on how I'm thinking about things and what I'm learning and what I'm seeing. So I will use some of this as part of it. And of course, now that we have our side chats every two weeks, that's your opportunity to put me on the hot seat. So with no further ado, I'm going to move on, keep in mind this is developing ideas, right? And I reserve my right to change my mind. And if I fall asleep standing up, that's it. Just pardon me for that too because I haven't been sleeping a lot. So I don't start with, I don't come in with solutions. And you already all probably have noticed that. I come in with a lot of questions and I come in with ideas. But those of you who have met me in our individual sessions by now know that I want to hear from you more than I want to be talking at you. And once I hear from you, I'll probably have some ideas and suggestions. So I've been asking you a lot of questions and I've been listening a lot. And one of the questions that I ask you is why are you here? Who are you here for? And what do you do for them every day, every morning that you wake up? In other words, in a more straightforward way, who's your boss? Anybody? Who do you think your boss is? Okay, one answer. Employees of the foundation, any more? The Rearers, okay. Humanity, very nice. Well, the positive news is that it's not me apparently, so I am abrogating your sensibility. Yay, okay. So your boss and mine is the user, okay? I want all of you guys to remember that. We're here because of the people who use the site and the people who contribute the knowledge to it. Okay, every day we wake up. Every night we go to sleep. These are the people we work for. These are the people I work for. Okay, who are they? Oh, this is a combined impression of what I can tell from our current numbers. About half a billion readers and about 70,000 editors. This means our job is to understand them. And it's really hard, obviously, to understand them in a one-on-one conversation. So we'll talk a little bit about what this understanding means. So let's see about what is ultimately our job. Our job is to make their experience the best experience it can be. And I don't just mean, like I was mentioning earlier, I don't just mean the software, right? I mean the experience overall. I mean the experience of them feeling motivated, them feeling engaged, them feeling loved for whatever, whichever way they will feel that way. And as you know, thanks to Eric, I made my first edit the other day. And I have some sense of what that experience is. But I want to make sure that all of us here start to really focus and understand that. And to tell you the truth, when I look at our numbers, they're telling a pretty grim story, right? Both of the, what this called, this is the slope for those of you who are not from the Sun's background, are showing the derivative, what's called the derivative. And they're both going in the wrong direction. Now there, we can debate whether the data is clean or whether it's right, but directionally, something is wrong here. So it's really important for us to recombine our united focus, both from the people here within this room and people who are out there, for our community on the users of our projects. So, you know, I've published a few emails out there and I posted a blog. And people told me, you know, it's inspirational. I think inspiration is important because that's why we're here. We're aspired to do the great thing for the good thing for the world. But in reality, things that we're doing here are very much operational things, right? And focus and execution are very important to us. So I want us to start thinking about that. And it all starts with measurement. So it all starts with understanding because we have so many users and we can't logistically interview every single one of them, right? At least not right away and ask them all of the smart questions and understand everyone individually. We need to understand their signals, their actions, their patterns of behavior. We need to measure everything that we do. And that's something that the executive team and the rest of you are going to start working on doing in a little bit more pragmatic way. And we need to learn from the data that we see, right? Measurements can be seen in different ways. They need to be interpreted, right? So all of us here need to start measuring and understanding what our users are actually doing and what are they motivated by. The next part, operational part, is ownership. We all have to own part of that experience and part of that behavior. Ownership is extremely important. If you don't own, you can't make a difference. And then you act. And finally, each and every one of us, our community and me are going to hold ourselves and each other accountable for what we own and what we act on. With that, it's really important that we get common goals across all of the teams here. We're all doing spectacular work in our own individual ways. But it's really important that we all focus on the common goal because when we unite and when we bring our ideas and our thoughts and our work together, we're that much more powerful. So if engineering team delivers a feature and the fundraising team and the grant-making team and the communication team are there to support them and all of us are working towards the same thing, the impact is going to be much larger. We measure the objectives that we have. Again, back to holding each other accountable to the deliverables and sometimes we'll fail and that's okay. We just need to fail fast and cut the things that we don't think are going to work. This is an important portion, right? Failure is part of success and I want all of us to embrace it here. Finally, we'll build things that are repeatable and we grow them. And we measure, we go back to measuring the improvement. But remember, we start with the measurement. We go through the measurement and then we measure at the end. That's how we know what we're good at and what we're not. So my responsibility here, my job for you to do for you is to give us all clarity of that operational purpose. To help us with our ownership, to hold myself and your job is to hold me accountable just as my job is to hold you accountable. What's yours? To think about how you can measure your success. To speak once you have your ideas and speak openly and don't hold back. And finally, the most important thing, your job is to drive. We're not just here to talk to each other. We're here on a mission, a very important one. So we're here to drive it. And with that, remember who we're here for. I wouldn't be here if I weren't here for this person. And that's my search appeal. I think I have, am I on time? Okay, so I don't think, I think we're pretty short on time today. I will ask that next time we'll have more Q&A time. So I want to make sure that we actually go through the agenda today. For those of you who are here, you can obviously talk to me tomorrow morning. We have a whole hour for that. And for those of you who are not, please go to my user page. I'm actually monitoring and responding to questions there. Okay. Thank you, everybody. Eric. Good morning, everyone. So we have a very full agenda as Laila mentioned. So I'm not going to spend a ton of time on the top level metrics, but I just want to call out one thing that we're observing right now, which is that there's a discrepancy that's growing between the numbers that we see in comscore and our own data. And the comscore numbers are actually going down, and our own data is holding steady when it comes to page-U numbers. So the page-U numbers, as you can see here, if you look at the all Wikimedia projects numbers, are relatively stable over the last months. In fact, we're not seeing quite the same pattern that we saw last year when it comes to PBs actually trending downwards for a while, whereas the comscore numbers have gone down to 465 million uniques. The comscore numbers are based on panel-based research, which means you identify a bunch of users who track their internet habits much like Nielsen ratings, and you try to come up with an estimate. So they're a little bit BS because they don't actually measure the reality. These numbers also, in fact, don't account for mobile usage and other usage patterns. So they're pretty shaky numbers to begin with. So we have more confidence in our own data of what we're registering on our servers. With that said, we're going to talk to comscore and figure out whether there is a trend here that we should be aware of. There's also a discrepancy between how comscore measures different sites. Some of them use what's called unified measurement, where sites actually embed beacon code from comscore into their sites to track their users. We don't do that. Increasingly, the top ten websites have that code embedded. We don't. That increases the delta between the quality of our measurement and theirs as well. So the mid to long-term plan will be to have our own UNIX measure that we can derive from our own data. But for now, our focus is to actually validate the reliability of this pastry data. So we really have confidence that the general traffic pattern that we're seeing here is correct. I wanted to give you a super quick window into a very different project which you may or may not have followed at all. So there's a project called Wiki Loves Earth which was initiated by originally Wikimedia Ukraine and has now been taken on by a few other chapter organizations and loose groups of volunteers around the world. And it's a pretty cool project that tries to document natural heritage. It's very much inspired by Wiki Loves Monuments and the difference here is that instead of going into your street and trying to find a building that's notable for some reason, you go into a wildlife reserve and you document nature. You document planet Earth. And what's interesting here so far is that the uptake has been fairly random in terms of which countries have decided to participate. So you've got an interesting mix of countries because it's basically been whoever signed up. These are the ones that are actually done. These are the ones that completed through May. So you've got Austria and Brazil. There's no consistency here in who signed up and for what reasons it all depended on like a critical mass of volunteers in a given country just coming together and doing it. The total of pictures uploaded right now is about 60,000 and they have their own little tracking tool at the end that you can see here. And you can see that Ukraine still by far is one of the more successful participants but Brazil was doing very well as well. They have almost 7,000 photos that have come through in their contest. The other thing that they're tracking which I think is cool is the actual usage of images which is another measure of success for these kinds of contests. So you can see that the usage varies actually pretty significantly from country to country anywhere from less than 1% to 10% to 15%. Of course, if you only have 77 images, that's not a very significant number. In the case of Brazil, the actual usage is still very, very low. So you've got a lot of people sending pictures in but not a lot of people adding them to Wikipedia articles. So this gives us a good sense of how we can help improve these pipelines. We've had in the past some success in just giving the community better tools to organize these kinds of contests. So you can see here that these links that go to campaigns actually give you a dark window into the dedicated upload pages for each country. So if you go to any one of them, you can immediately upload pictures that are then correctly categorized in the right place and can then be discovered here. But there's a lot more that we could do. Generally speaking, it's not as large-scale as Wikilocks Monuments. It might never be because it's a more specialized photography contest but it's still a cool thing to keep an eye on and it's going to help bring in lots of new users. With that in mind, I want to turn over to Julius Conner, who's going to give us an update on the visual editor. Mobile team has been doing a lot of work porting visual editor over to tablets. So he's going to show the current experience. Hi. So currently... Well, this is going to be difficult. I'm going to make it a little larger. We'll see. Let's switch to... Okay, we'll see. Wow. I can drive. Yeah, okay. I clicked it. So presently we have visual editor in the so-called experimental mode, which not too many people use and it's good that not too many people use it because it's experimental and breaks often. Soon we will be moving it to beta mode. Originally we planned to move it to stable mode. So for all the users, at the same time we're redirecting tablet users by default to the mobile website, which is June 17th. But we decided that we don't want to rush things and we want to work on it a little bit more, fix all the bugs. So to get the visual editor right now you need to first enable beta mode, then again go to settings, enable experimental mode. So it's pretty well hidden out there. And then assuming that you are using some device with a screen that's big enough and you're logged in actually, then by default, just give me a second, this is just my local DevWiki, I don't have one character password in production. So first of all you can notice that this is actually bigger than a mobile phone screen and we made some changes to typography, not so long ago, so that actually everything is wider and looks better on a tablet. So if you press the edit icon, it loads the visual editor. And if you really don't like visual editor, we have an editor switcher which lets you go to the source editor, which is the old editor we had. So just a quick reminder of how it worked. You can make some changes, you go here, you get a preview, and you can save your changes. So this is the old editor. Now the new one, I mean the new one, the new one on tablets, is a simplified version of what we have on desktop. So you can edit and see the changes, what you see is what you get basically. So you can do basic text formatting. Now also one important aspect is adding links to articles, so we have a link inspector. It looks kind of different than on desktop because we had to change the appearance because basically tablets have very often those additional controls. If you select text, something pops up and if something from the system pops up and something from the ePops up, there's just too many pop-ups. So let's say I will link to something. Now if I go back, this is already a link, and something we've been working on recently are references. So you can add certain types of references. Website is an easy one because there are just two fields to fill in. And this is also simplified. If you use that on desktop, you would know that the interface is kind of... Okay, thank you. That's helpful. So the interface is simplified. We only provide the default fields. We don't want to overwhelm people who are using V for the first time on tablets with too many options. Okay, let me just do this. So now we just added a reference. If we go down, we have a reference section. We have a new reference here. Okay. And we can save. Now obviously there's no preview because the changes you do, that's what you will get. So you can type a summary. You don't have to. You can save. Now I added an external link, so I have to type... World world? Okay, that's good. And there we go. And it's saved. The link works. Goes to another page. Oh my. And the reference has been added. Yeah. So it's pretty basic, but it works. Some features like the link inspector kind of break on some platforms like iOS, which is 90% of our users. That's why we're delaying putting this to stable. But it's almost ready. Do you have any questions? Why is it tablet only? It seems like there's almost enough room to do that on an iPhone from what you showed. Yes. In future we will also enable it for phones, I think. I think the main reason right now is performance. Tablets in general have a bit of a better specifications than phones. They're faster. And improving these performance is an ongoing work. So both in the VE team and in our team. So at first we don't want to launch it everywhere. We just want to launch it on tablets. Also one more reason is that currently tablets go to the desktop site which has VE enabled so they can use it. So when we switch tablets to the mobile site we want to give them the same functionality. While mobile phone users are used to not having VE. Awesome work by the mobile web team and the VE team. I want to highlight that as we go into the next fiscal year and work with all our teams on their planning and other factors that we're emphasizing is really getting every team within engineering and product that builds user-facing functionality to be a multi-device team from the start. That's not something that we can just put on teams mandate and say, oh, now you're a multi-device team. We have to give them the training, the resources and the support and the shared libraries, the shared code base necessary to make that possible. So there's a lot of work that is happening now in the libraries to standardize templates to give teams more support to be multi-device ready that in some cases is going to actually slow down the velocity of delivering functionality for the desktop. It has to in order to make that possible but it's a necessary transition for us as an organization. And this initial port of visual editor to tablets is just the start. Really every bit of user functionality that we build has to be ready to be used across devices, across capabilities. So with that said I want to turn over to the next speaker and that would be Andre who is giving a remote update on the migration that is underway of first the bug tracking system and then other components of our tooling to fabricator. Andre. Hello everybody. I'd like to share my screen. Let me try that. All right. So I hope you see my screen now. If you haven't heard of fabricator let me start with describing the problem we'd like to solve. And Wikimedia there's several tools used currently for infrastructure handling. So for example, the operations team is mostly using RT request tracker for their tasks. Product management planning is done, for example, in Mingdal. Some other teams use Trello. Wikidata and also analytics are using Scrum Box on top of our bug tracker. I only found this screenshot because the instance somehow is down currently. This is on top of Buxilla which is probably the most known one which is our issue tracking system. This is how it looks like. And for example in Buxilla if I scroll down you see there we have comments if there is a new patch available for this bug report which links to Garrett which is our code review system and it looks like this. So as you can see discussion on issues is mostly scattered and it's pretty hard to follow. So for the past months, Guillaume and I have been looking at other tools available. We collected and evaluated the different needs of different development teams and we ran an RFC to gather community feedback and the result is that we are working towards moving all this into one tool which is called Fabricator. So here currently you see our test instance on Wikimedia Labs which is on fab.wmflabs.org and we also use this test instance for planning the actual immigration. There is no time frame set currently. To give you a quick idea I just want to show you three pages in there. The planning we are doing is in a project called Wikimedia Fabricator Day 1. So here you can see the open tasks that we still have to tackle or sort out. We have a backlog but you also see stuff that still needs discussion or more input tasks that can be picked up that are ready to go. Some stuff we are waiting for upstream feedback from the upstream developers and doing is what is currently being worked on. Fabricator as you can see here on the left is a set of suite of tools. So for example you also have bug tracking which looks like this and for example compared to Buxilla you can already see that it hides some complexity when it comes to the fields. So if you want to change a field you have to click at a task which we don't target for the very first day of production but it also includes for example a code review tool which you can see here with a test commit. So the current plan you can find on mediawiki.org wiki.com. There is the migration plan. So the current focus and the current team this is all still might be subject to change is Mukundu model currently working on authentication and a single user login and also upstream working on access restrictions for some projects because that's what we currently have in RT and Buxilla for example tickets that are about security vulnerabilities. Chase pattern of ops currently looking into setting up production instance of fabricator and we currently investigate deploying a trusted user tool which means we can track or manage users in the wiki media sphere who have specific access to things. So as you can see there is more steps afterwards for example we still have to investigate how to migrate the data from Buxilla into the system. We do not plan to migrate Garrett the code review tool currently and Jenkins the automated tested continues integration tool for day one but we want to do this later and to continue with the people involved in this James Forester is our proxy for the product management needs and Greg Rosemeyer for the platform needs and of course there is way more people involved to comment and help like Angel and many others but I shouldn't mention now because we are running out of time. For more information and also for videos go to media wiki.org there is the fabricator website it links to the board for day one planning that you just saw and get involved. There is a get involved section. Help is always welcome if you write code or just providing input and discussions or just asking questions. Thank you. Thank you Andre. Any questions for Andre? Just a 10 second test. Who has an account in the lab? Is that a fabricator? Can you raise your hands? Okay. All the rest you can also register and try it out because this is good for software projects but this is also good for any kind of projects. Actually the legal is not a software project. The trusted tool for LCA. Give it a try. I want to note that fabricator is a tool that we looked at when we actually adopted Garrett as a code review tool and at that time more than a year ago now it wasn't quite ready to be the be all and all solution for re-committee engineering yet but the upstream behind fabricator has been doing amazing work over the last year and beyond to really make this a usable solution for many many different use cases. Maybe a little bit too many if you look at all the different projects under the fabricator umbrella but Evan and the fabricator team are incredible to work with. Evan Priestley is the founder of that project. He was at the office yesterday. He's in the fabricator labs instance and helping us and answering questions. It's just been an amazing partnership so far. The tool itself is written in PHP which is Java in GWT which makes it a lot easier for us to develop and customize as we need to. I think it's the right path forward but we're going to take it very carefully because it's going to affect a lot of workflows and I want to thank you Andre, Kim, Chase, Mukunda, everyone else who's been part of this process so far. I can see your activity on the labs instance ticking off boxes. It's just wonderful to see the energy that you've been putting into it. Thank you so much. Next up we have a quick update from Steven Walling, Product Manager for the growth team on the experiment that they ran trying to get anonymous users and Wikipedia to create accounts. Hi everybody, I'm Steven like Eric said and also on the growth team are Moise and Katie who are here in the office from the UX team. Erin Hafaker who's in Minneapolis from the research and data team so thank you. On the engineering side we have Sam Smith in London, Matt Flaschen in Philly and New Rob Mowen in Portland who joined us from VE team. We're going to talk about inviting anonymous editors to register. Why would we do this? We do this for two reasons. The team's goal is to work on growing the total active editors and that's a measurement of our core community and it only includes registered users but more importantly we think if you're the kind of person who wants to edit Wikipedia repeatedly you have a better experience if you log in even though you can edit anonymously you have an identity on the site you can get notifications, track your contributions, a whole host of other things. So we're really doing this because we think if you're the kind of person who's a real born and bred Wikipedia you want to log in. The other reason is if we're looking at growing the community there's a lot of unregistered on the editing every month which is extremely large if you compare that to the community of 70,000 active registered editors who are making 5 plus a month. And so our question is can we just ask people to sign up? This isn't actually a given. It's entirely possible or at least a reasonable theory that a large proportion of the people who edit anonymously are fully aware of the tradeoffs between being anonymous and being registered. They maybe even had accounts in the past and they're not interested anymore because they got yelled at or reverted like we could just not be able to ask people to sign up. And the current legacy experience actually does have some hints in it that ask people to sign up other than the create account and login links that are present on every single page in the top right. When you edit anonymously we actually tell you hey you're not logged in and there are benefits to doing this. This forms a small but significant proportion of the people who are signing up today about 12% on English Wikipedia and smaller or larger on some non-English Wikipedia's. But we wanted to test two new ideas for asking people to sign up and we actually ran an A.B. test for a week in English, German, French and Italian which is even if the test was a failure it was pretty exciting because we've never run simultaneous A.B. tests before across multiple languages. So it's interesting to compare some of the data from these projects. The first version we tested we're asking users to sign up before they actually click it. I know I'm confusing you because I have the screen size in German but what the workflow looks like is I'm an anonymous editor as soon as I click edit either at the page level or the section level I'm asked to sign up and edit Wikipedia it's free easy and it just takes a moment and we don't collect any personal information. And then if I see you say no thanks or I click edit again then I can continue to edit anonymously. Or I get sent to the registration page after which after I sign up I actually get sent back to the same article in edit mode so that I'm not interrupting the workflow as much as possible. The second version that we tested simultaneously against that but each user only gets one version is post edit. I've just completed an edit anonymously and afterwards we point to the create account function and say hey you should sign up for Wikipedia here's all the things that you'll get as part of signing up a list of like five reasons and then that sends people to the sign up page and then they actually get sent after that to the article they were reading before and they get the normal onboarding workflow. So results we look at the results of this test on like three basic levels one is how many additional registrations did we get to how many of those people were activated as in they made at least one or more edits. And then the third is how many of them became active editors that made five or more edits to articles that would be counted as part of our total active editors metric. So for sign up results so the percentage of people who actually registered out of the whole group of anonymous editors who were bucketed in that experiment who at least clicked edit during the test we actually gained eleven thousand sign ups and then an additional thousand sign ups in the post edit version. So if you consider like two or three hundred thousand anonymous people editing every month a change from one percent of people converting to sign ups after or during editing to three percent is a pretty big difference in registrations. The second rate looking at how many of these people who signed up actually edited in each condition the really the really good news is that the test for editors in who were anonymous were either the same or better than people who were natural sign ups who didn't get any experience. Now this is kind of miraculous considering usually whenever we do something that drastically increases new sign ups we're gaining people who need more hand holding or aren't as likely to edit and you see a consummate decrease in the number of the rate of people who actually make one or more edits to articles. One of the theories about why this is is because the users in the pre edit condition are really highly motivated to edit. They just clicked edit we say sign up to edit Wikipedia then we send them back there so it's really likely that they would do so and then in terms of actually how many of those people went on and became active editors who'd be counted as in our total count of the community in the post edit version there's a statistically significant increase so they're actually better than control or pre edit and the pre edit version is definitely not worse so what that means and for non English results I didn't want to dig into them too much because they're four other Wikis and we only have so much time but the conversion rates are roughly the same and then the scale of how many people are actually signing up and editing is significantly smaller so we're talking thousands in English and hundreds or tens in other language Wikipedia's so like we got more people to register and they are at least as likely to edit if not better holy crap that's great news but not so fast because there's a few caveats one is that these conversion rates might fall over time and then reach a new baseline this is the first time we run a site-wide test where we take a huge proportion of the people who are editing anonymously and ask them to register so we're sort of like front loading by catching everybody who's never been asked really prominently to register and then getting them to sign up so it might fall over time the other one is we didn't we haven't yet measured but we're going to the impact on IP editor so the people who said no thanks I just want to edit as an IP keep going we are interrupting those people so we want to make sure that we didn't end up causing a decrease in their editing activity alongside that and then three during this test we showed each of those versions multiple times to the same user which is a nagging thing we did measure how many multiple impressions each of those users had so we need to figure out how many times did it take to ask people to sign up for each individual in order to actually get them to do that and there'll be a report on meta if you go to meta and look up growth team there's a list of all of our research and data so questions questions from IRC do we know why people choose to edit anonymous like actively choose as opposed to just lacks into not having account so I don't think we know in the sense that we've measured the proportions between these different things but we have basic ideas of all of the like personas of people who edit anonymously there are logged out Wikipedia and there are people who've edited anonymously repeatedly over the time and then there's actually making their first edit to the site it's some mix of that and we're going to dig into the data from the experiment more to figure that out because for the two weeks that we tracked anonymous editing behavior it'll actually give us the first real like data driven look at how many of those people were registered before they went out and logged out and edited again etc etc so this is really great thank you how long until we have a full picture like what you're talking about with all of their precautions and downstream effects like a week or two I mean ask Aaron but the man is the most productive human being I've ever worked with so like Jared since it seemed like the pre edit and the post edit had different positive effects one was about creating sign ups and one was about actually retaining editors was there any users who saw both and have you thought about what possible effect that would have other than possibly being a little bit annoying and naggy no the bucketing work such that they're exclusive to each other they could not get both versions and I tested that but did you have any theories about what possible effect that could have why that is what would happen if someone said no to the first oh if we do the pre edit and then we do edit on top of that I have no idea we could test it potentially asking them to log in or register on safe right so in between the reason I'm saying bringing this up because that's my problem I just forget to log in yeah testing against to sign up and say workflow is something we definitely would potentially do especially in visual editor I don't think and Marianna and Dan could correct me it's going to be part of the first release but on apps it's actually significantly easier technically for us to do that sign up and save and then retain the person's edit and actually save it for them after they log in or sign up so I think we're actually going to test that on mobile first so yeah that's really awesome and other question that is related to what we're seeing on mobile so we know that mobile the most effective way of getting people to create accounts and actually be activated is by actually intercepting their desire to edit so in a way mobile is very different because you need to register an account as of today to be able to make edits do you feel the direction this taking especially a direction in which we'll be strongly encouraging editors to register accounts and make it very prominent in the final product to do so in order to have basically increased number of registered accounts yes though not necessarily for that reason so I think even if the test had been a failure one of the things that we learn out of this that's important is when we talk about the editor community how much of that is anonymous so we don't count them in total active editors partially because it's just really hard to measure that without giving people a bunch of cookies and decreasing performance and causing problems so we'll get a sense of like how much active anonymous editors really contribute and talk about things like should we be tracking that as part of our top line metrics alongside registered users that kind of thing but on the other hand to honestly answer your question like if this works without decreasing the editing behavior of people who want to edit anonymously and we get more additional registered contributors the answer is yes we're probably going to strongly encourage people to do that we think it's better for them thank you Stephen go to move on and the next update is from the multimedia team on the work that they're planning on the upload wizard the tool that's used on Wikipedia comments to upload media and I also just want to take a brief second to congratulate Fabrice and the media team on the launch of media viewer which is now running on English Wikipedia German Wikipedia and the Germans have not revolted yet so it's an amazing new user experience for pictures so if you haven't tried it yet click a photo in any of our projects and give it a spin and now about contributions so we're happy that we've published the first part of the mic mic mic happy that we've accomplished the first part of the multimedia vision we showed you in January which is to help readers get a better viewer experience so they help all users enjoy media in the way that should be seen we're now shifting our focus to address the needs of contributors and upload wizard is a really important project that will be our main user facing project in the coming year the goal is going to be to engage more users to contribute and also to help people add more useful media to articles and in the process we plan to provide a much smoother experience that exists today and we have a lot of bugs to fix and a lot of technology debt to address so the goal for the experience is covers a wide range of objectives from learning how to upload to being able to add information anytime during the upload to eventually being able to upload anywhere being able to upload from visual editor flow so Pao is going to show us our vision for how we might improve this experience in coming years Pao Hi, as Fabriz said I wanted to show you some design patterns that we think that can help make the uploading experience more fluent I illustrate those in an example scenario where I show some mock-ups that should not be taken literally they are just intended to provide the general idea but for example when I try to upload a new picture it would be great to have a general idea of what means to contribute an image to comments for example as opposed to another social need work which are the implications of the single glance and also get me encouraged to just drag and drop it so I can start the process faster while I am uploading the picture I can really see the progress in real time but not only that also from the very first moment it's interesting to be able to provide the details of that image to describe it just from the start of the upload that not only saves me time but also puts me in a more relaxed situation where I know things are on track and I can spend more time doing a better description when doing that description of the image it's also interesting to apply different design patterns that are common on web form design to use small defaults to allow users to reuse information that they used before so that we don't make them do the same work again and again and the same happens when propagating information across multiple uploads one specific pain point of the current process that many users complain about is categorization it's important but it's hard and I think that we can aid to do the same help them use recent and previous categories they have been using explore the subcategory hierarchy and that will make for better categorizations an easy and fast categorization process as I mentioned dealing with multiple files it also adds a lot of complexity and if I don't want to add just one picture but the whole collection of images of animals that I'm showing here it's really helpful if I can select multiple files, organize them provide information to several files at a time and not only to contribute new information it's useful but also to use it in many different places so I think that it's also important to encourage easy ways to reuse this information simplify instructions, make a single link work everywhere and not only that but also other projects that deal with media like I'm editing on visual editor I want to insert an image it's really great if I can be I can have access to contribute new images from there or I can or it can anticipate my needs and show me my recent uploads so that I can just insert the images that I just uploaded some seconds ago another another interesting aspect is to try to avoid this to be an upfront effort only so that I provide information only while I upload things but also to use the same patterns and the same aids when I edit existing images that are already uploaded or not to force users to provide all even the required information at the very beginning to be able to edit things at different points in time great thank you pal so this is a great vision we're really excited to start working on it it will take time to do it it's not going to all get done this year but for this year for the first three quarters we're basically looking at progressively developing some of these features as well as improving the platform and then working on some special projects so in the first quarter we'll implement things like the progress bar and better help we'll fix a lot more bugs we'll take on shock uploads which basically is related to uploading large files and most importantly start getting a whole bunch of metrics going so we can really see where we're going and the Q2 we'll start working with wiki data to implement structured data on commons which will make it possible for us to do things like adding information anytime during the upload as well as dealing with some other features like the category tool which would depend on this and then Q3 in the winter we'll start working with the visual editor and the flow teams to make this upload wizard technology available for their tools and their users as well as address things like multiple files and cross wiki upload so it's a big agenda but we're very jazzed about taking on this new initiative and stay tuned for a new and upgraded upload wizard coming soon to a wizard to his screen near you thank you I don't think we have time for questions since we're already over quick question have you had any interest from professional photographers use tools like lightroom or aperture to have a plug in to do an upload through that flow that's how I tend to do most of my photo processing on my desktop and then upload it to other third party sites there's certainly some interest in doing this but we're probably going to have to first cover the first basic steps and it's probably a year or two initiative but your point is well taken thank you Fabrice thank you Pao we have two quick updates from grantmaking just a quick note that we are running over so you guys could keep one of them a little bit shorter alright so keep it quick was that the message alright well it's really nice to see you all again I've been gone for a few months so I really am glad to see you all here again and to tell you a little bit about this one of the four programs that we have to support are different constituents in the movement and that is the annual plan grants which is run by the funds dissemination committee the FDC do I have the clicker so the funds dissemination committee met recently and many of you I'm sure have followed their discussions and their recommendations to the board but you can see they deliberated on three proposals one was from the center for internet and society in India one from Wikimedia Norway and another from Wikimedia France and they also reviewed the first version of the annual plan of the WMF which was the first time they have done that so you can see here a little bit about the recommendations that were made and how they compare against the requests that came in from the organization as well as the comparison between last year's grant and this year's grant so you do see that there's a little bit between the percentages year over year but we didn't want to talk too much about that and rather tell you a little bit about some of the proposals that are being funded because they're really interesting we have CIS in India that is doing some really innovative and interesting work around different language communities in India they are working specifically with one or two or a few targeted language groups working on different outreach and community gatherings to increase some articles created as well as to bolster and support the community we have Wikimedia Norway who is doing a lot of GLAM partnership work in order to produce high quality photos they're doing trainings workshops and editathons with different GLAM professionals in Norway and we have Wikimedia France who is doing some really innovative work around evaluation of how offline works offline work like workshops and community gatherings can support online work because one of the things we're most interested in is how is all this offline work contributing to the projects and that's a big question that we are always working to answer we had some learning from this particular round of the FDC's deliberation we're starting to see some shifts and impact strategy that are focusing on this offline to online shift we saw that particularly highlighted well by Wikimedia France we're also seeing growing depth of self-defined metrics and indicators but we need more we're starting to do some analysis on the first round of what we call impact reports that are coming in to us to really understand how organizations are tracking the impact that they're having online and so we're seeing more and more willingness and contribution by these grantees in defining their own metrics how are they hoping to grow the communities and support the online projects then we saw the FDC the style and substance was universally appreciated so far because there was a lot of detail a lot of the grantees have said we're interested in how you got to this particular number it may seem kind of random if you don't understand the FDC's deliberation and thinking so they did provide a lot more and finally WMF all of us here were a part of the review of the WMF's annual plan this was the first time that the WMF participated as a part of the FDC process in this way and so this was a really interesting experience for the FDC they did not make a recommendation or a grant allocation to the WMF and I think there was a lot of learning that we'll see transforming the process for the next round of how the WMF participates we are seeking four new members to join the funds dissemination committee so this is where I ask all of you to help get the word out and to spread the news you can see what fun we have I think Anasuya is really enjoying herself in this picture we have two of our community members and we're supported by two of the members of the board Patricio and Bishaka we have several folks in this office who support the team too and you'll see all of our friendly faces here we're looking specifically for people who have backgrounds in managing projects finance, grant making of course wikimedia experience online and offline and a diversity of folks and perspectives because what makes the process so unique and interesting is really the diversity we are taking nominations through June 15 now I will pass it over to Jesse for the last two minutes I'll go really quickly so part of this immediately following the deliberation of deciding where the grant funds were going was a reconvening of the FDC advisory group quick overview we as the grant makers aren't trying to be like a lot how a lot of other grant makers are very top down community driven grant making process in conjunction with our site in all of our values as part of that there was an advisory group that helped establish the FDC two years ago to begin with and part of the framework behind it was that after two years of the FDC running they would reconvene and assess what had happened and make a recommendation to our executive director as to whether or not the program should continue and with what modifications if any so anyway we had this convening a week or two ago and they gave a thumbs up that we should continue and they'll have some improvement points in there detailed recommendation will be published publicly as part of that I'm Jesse I lead the learning and evaluation work for the grant making team and we helped provide a lot of information to the advisory group to think about what is the impact of the six million dollars that this group is giving out annually and there's a more detailed report that we have that we're doing a few webinars this month but we've been trying to look at what's the online impact of the work that has gone on what are the users actually contributing through a lot of the work our grantees are doing and we have a lot of interesting data and we'll do a lot more to share this with you hopefully the July metrics month will do a bit more to pass this off that's good thank you thanks Eric I have the pleasure of cutting into lunch hi everybody I'm going to talk about individual engagement grants this is our third round of grants and we've got a bunch of new projects just a reminder these grants support small teams or individuals to run projects six months experiments aimed at having online impact and we're really looking for more readers more editors more content so this just kind of gives you an idea of the overall shape of what we're funding this round we're funding 12 projects which is more than ever before the dollar amount is about the same as in past rounds I think we've seen the quality of proposals improve over time thanks to some process iterations we've made projects are ranging from $600 to $22,600 so there's a big range and you can see we're funding a lot of tools a fair amount of online community organizing and then a little bit of research and some of the more kind of standard outreach initiatives just to give you a sense of the spread our grantees are very global so we're funding folks in 10 different countries and on four different wiki media projects and then multiple languages within those projects and particularly interesting for us this year it's our first time funding anything with Wiktionary or wiki voyage so it'll be interesting to see sort of what happens with those communities as a result of those projects so I'm not going to talk about all 12 grants but I'm going to walk you through a few of the sort of what's particularly new and particularly experimental in this round one first theme that I'm seeing is mobile so we're funding two different projects that are working on mobile apps this time one is the wiki quiz it's some Chinese wikipedia Chinese wikipedia readership is quite low there are other government and Chinese encyclopedias and so raising just readership and awareness of wikipedia in China is a big challenge for that community so they're building a game a quiz app and to see what that does for readership Addis likes to talk about it as wikipedia addiction the other one which you've got screenshot here is an app called wiki track and I know our mobile team is interested in sort of following along what they learn from this it's really an app for power users it's editors who are tracking recent changes right watch lists sort of all of those things on the go which are mobile apps are not at least yet focused on this app already exists in some of the Indian languages and has thousands of users so the idea is to consolidate it and make it available for all languages and not just wikipedia but other projects as well the next one is a mentorship proposal so some of you may remember a couple of years ago we supported the community to build a project called tea house and tea house was basically you know creating a new experience for mentorship that is a many to many mentorship space which you still find on english wikipedia today we're hoping with this project we're going to do the same thing for one on one mentorship so this logo here is I think the logo that is still used in the current adopt a user program on english wikipedia and it was created in 2005 so there's some room for the experience to be updated and you know the other thing with that current program is it's like a six month commitment you know you sort of sign on to a mentor and you're going to work yourself through curriculum it could last six months to a year which is not really how web based engagement works these days so the idea is to start to break down that learning into smaller bits and match people up for shorter term engagements and just start to build more connections between new users and long time editors so we're pretty excited to see where this one goes and then the last thing that is sort of new for us in this round is we're funding research for the first time we're funding two research projects one is looking at category systems paul is an english wikipedia who likes to you know say that english wikipedia are always having the same conversations about categories and sort of the same debates continue again and again and paul wants to provide some data into those conversations and to see if the community makes different or new decisions as a result of that and then we're also funding one project looking at the gender gap a lot of the data and research that we have particularly in terms of internal community research on the gender gap is from 2011 so amanda is conducting lots of interviews she'll be running a survey and just starting to help us put some more data to assess current efforts to close the gap and then for future avenues that we could be working in so that's it I would encourage you guys if you want to follow along these projects if you go to the IEG pages you can find all of the projects our grantees are always sort of submitting their work outwards so that everybody can comment and participate they're also blogging a lot so you know feel free to look for them on the blog and then of course come to idea lab share your ideas give people feedback on their ideas because we're always building the next round of experiments thank you siku questions for your secret or kitty any questions or other grant making stuff going okay time for lunch thank you everyone yes