Grants talk:APG/Proposals/2015-2016 round1/Wikimedia CH/Proposal form

From Meta, a Wikimedia project coordination wiki

Congratulations on a excellent plan![edit]

I have read through your plan and van only express my praise for a work well done. I find your programs, goals and targets all relevant, well thought through and clearly presented. I am extra pleased knowing that your history with FDC have been somewhat unsatisfactory (I was particpating in FDC in two of these rounds) and to see that you this time utilize your full experiences and competence and wish you a good productive 2016. And also that make your good plan/work known to others by presnting parts of it at appropraite conferences etc.Anders Wennersten (talk) 15:26, 7 October 2015 (UTC)Reply[reply]

Hey thanks Anders! Just to be clear, we'd like to thank Katie, Winifred and Jaime for their input: it really helped. Stephane (WMCH) (talk) 15:25, 9 October 2015 (UTC)Reply[reply]

Addendum / Correction : "Tools" program[edit]

I actually got the numbers wrong in the Kiwix part : we already reach 100,000 downloads/yr through the Android store only: the actual number for all platforms is actually closer to 500,000 / year.

The targets should therefore be updated as such:

Roll out on Raspberry pi (Q1); 2 mini-hackathons (Q2/Q4); 750,000 downloads across all platforms; engage at least with two major donors for 2017 (for a total of USD 100-150,000); renew existing partnerships with at least two of our current partners; extend reach to more than 1,000,000 individuals.

(the corresponding item 3 of table 3 should also be updated by one order of magnitude).

Also in the notes "Rolling out Kiwix on Raspberry pi should allow us to decrease our cost-per-unit by another 20%"

So eventhough the submission deadline is way past, we'd like to make to the above amendments.

Thanks, Stephane (WMCH) (talk) 15:25, 9 October 2015 (UTC)Reply[reply]

Hi, Stephane (WMCH), please go ahead and update it with the strikethrough text. Thank you. KLove (WMF) (talk) 22:28, 13 October 2015 (UTC)Reply[reply]
Done, thanks. Stephane (WMCH) (talk) 06:07, 14 October 2015 (UTC)Reply[reply]

questions from FDC staff[edit]

Hi Stephane and colleagues at WMCH - thank you for submitting this annual plan grant proposal. It's been reviewed by the FDC staff, and we have a few questions for you. Let me know if any of them is not unclear. Many thanks! KLove (WMF) (talk) 22:29, 13 October 2015 (UTC)Reply[reply]

Program and organization questions[edit]

  • Under the Tools program you note you intend to do a gap analysis, followed by beta-launching and sharing of tools. What kind of tools are you thinking about? Does your current staff have the required expertise to engage in this work?
    • Considering the budget we're looking into (10-20k) it's definitely not some sort of GLAM-wikitoolset II (though in the medium term we intend to pick that one up again). We aim much lower and think of tools like variants of Baglama or Coord. Basically what we need is not something fundamentally new and research-intensive, but something with a better granularity and more reliable: no need to reinvent the wheel. While we do not have the staffing to do this development internally (we'll outsource to partners) we do however, have the required internal expertise for managing such projects. The approach will be ITIL compliant so as to start from a service design and strategy and evaluate steps as we move along: one of our staff has 20 years of experience in IT projects and 10 years of experience in IT governance, three certifications in ITIL and COBIT and is planning a certification in PMI next month... The idea is to make good use of what we already know and let others do what we cannot do.

See also our response to Risker below for a more detailed answer.

  • We’d like to know more about the process you intend to build to monitor accessibility of Wikisource. Can you explain where the interest for that project is coming from and the specific role the chapter will play?
    • The idea came up after a speech of an engineer with low vision done during a meeting of Frontenders explaining the difficulties and the barriers to use internet. Here is more or less what he said:
    1. "Do not imagine that an integration of a TTL (Text-to-speech) or of a voice-to-text can facilitate blind people. Everyone *already* has their own tools to surf internet, they select and learn the use of these tools which are essential for a good browsing experience. The configuration and the settings of the TTL or of a voice-to-text are also very personal and follow the individual needs. For a person with normal vision the speed of reading of these tools is too high, but for blind people this is normal, for this reason this is an additional point to take care to avoid to propose a different TTL;
    2. The tools and the desktop configurations are set at the operating system and they are general, so the main need is to be a "standard" in order to be easily compatible with these settings;
    3. We discovered that Wikimedia projects are really appreciated because "well-structured". Some HTML tags like paragraphs or headers and some elements like the table of contents which are for us graphical elements, for blind people they are really essential for a good surfing;
    4. But there are also some elements which can turn a page in an unusable page or are barriers to usage, for instance some templates and captcha. Captcha is a real blocking element for blind people;
    5. Wikisource is the best candidate for this project because it's already well-formed to generate good e-books, we are sure that some guidelines and checking can give good benefits also to the generation of the e-books."
    Basically Wikisource is the most homogeneous Wikimedia project we have, and what we aim at is to collect feedback and work around the requirements suggested by the "real" final users. The role that Wikimedia CH would take is to help the community and to help specific associations for blind people to reduce the gap (which seems to be limited in our opinion) to solve this "diversity" but also to measure and monitor the distance of this gap. There should be good learning, and we see no reason why this couldn't be shared with other language communities at later stages.
  • In the proposal, you have noted that your objective is to bring several thousand high-quality images to Commons. You note your goal is to increase the pageviews of WMCH-supported media by 5-7%. What is the target number for those pageviews overall?
    • We currently have around 13,000,000 page views / month (on average 2015 YTD). 5-7% of this should increase the total by 650 to 900’000 extra page views. To be clear, importing those images to commons isn’t enough by and of itself: other elements in our programs (tool, workshops and edit-a-thons) should allow us to identify which articles have no or poor illustrations and then improve the related articles with the uploads that were made.
  • FDC staff and WMCH have discussed extensively the challenges of measuring the impact of Kiwix. We know it’s an offline tool that expands the reach of Wikipedia. Given that you’ve set your targets around events, donors, partners, and downloads, we’d like to understand more around how you will assess whether this project needs further extension or support. How are you considering usage at this point?
    • We haven’t been very good at tracking partnerships, to be honest, and some work is needed there: for instance, we found out that one of our partners (Ideas Box) is involved in the Connected Learning Initiative of the High Commissariat for Refugees. They have programs in Burundi, Jordan, Lebanon and other non-refugees sites, and recently won the Google Impact challenge which should allow them to roll out 1'000 containers over the next 3 years and potentially impact 5 million people. Whether they can do it or not, we want to make sure we can accompany them and others since, as far as we know, most offline wikipedia initiatives that circulate to this day seem to be running on Kiwix - we simply didn’t track them because we did not have resources to (but the HCR does!). So to answer your question, we considered usage to be a 1x multiplier of our number of downloads; it is as good a guess as saying that all people who download it may not use it, and those who do use it may do it in groups (because the whole thing is offline, we can’t set up usage data to be sent back to the mothership).
  • The fact that we’ve been invited to join Connected Learning is a good indication that our work is recognized, however, and that our reach is probably way beyond that (and can be physically measured by others on the field). Practically speaking, this will also allow us to get a better overview of who’s doing what, improve estimates of our actual reach (for instance, how many kiwix keys are being circulated to North Korea? How many kids are serviced through one Ideas Box? If we know whom to ask we should get more solid estimates), and better streamline feedback from distributors (we already get feedback from users through a dedicated channel, but that’s more along the lines of bug-fixing). Ideally, we should be able to have a somewhat comprehensive list of institutional partners using Kiwix, track whether they keep using it year-on-year, and based on their feedback improve the quality of our service. In the medium term, increasing our visibility in these forums should also help establish Kiwix plugs as an officially promoted go-to solution (as opposite to the de facto standard it already is.)
  • Can you tell us more about how the organization’s staff are being managed in the absence of an ED?
    • In general, staff is pretty independent and very professional in what they are doing: they started the year with a discretionary community support budget commensurate with their demographics and a set of objectives; this was carried out without much trouble - they’re good people, reacted as a team when the CSO left to keep their common activities going (e.g. staff meetings, coordinating newsletters, etc.) and we are lucky to have them. There’s been some overtime that will need catching up (or paying) soon enough, but that was to be expected. In the meantime, all official decision-making has been transferred to two board members who wrote the present proposal (and other reports), negotiated the renewal of our Fundraising agreement, and made the planning for next year, amongst others.
  • Can you give us a general idea of how much staff time you will spend on each program?
    • Only as a ballpark figure:
    DE CL: 0.2 OpenData; 0.6 Community; 0.1 GLAM; 0.1 admin.
    FR CL: 0.2 Community; 0.4 fundraising; 0.1 admin.
    IT CL: 0.1 Community; 0.2 GLAM/Outreach; 0.1 tools; 0.1 admin;
    Admin manager: 0.5 admin
    ED: Tbd.
    Note: Everyone is booked to attend and help Wikimania next June.

Budget questions[edit]

  • We will need a more detailed budget with expenses by line items. You may submit a copy of this privately, if you can’t share the details at that level publicly.
    • Will send you an itemized list shortly. Nothing secret but it is simpler that way.
    update: here's a link! [1]
  • How does this large amount of Wikimania funding fit in with the objectives of your community program? What exactly will this 25,000 CHF cover?
    • This « large amount of Wikimania funding » corresponds for the most part to staff time we intend to dedicate (or have started to dedicate) to the event: consider it a ballpark figure of having all hands on deck for 10 full days (as not everyone works full time, this includes overtime compensation for the 5 FTE it represents), plus work needed to secure Schengen visas for Global South attendees. We put the activity in the community program because its core demographic is made of active users (we might have made it a separate item but at the end of the day that seemed rather artificial).
  • What are the expenses you anticipate for the tools program?
    • Around 55kCHF for the whole thing (+ staff time). If you mean the "tools" within the "Tools" programs (a confusing choice of names, in hindsight), this is more between 10-20k, but there probably will be some pro bono from one of our partners (University of Applied Sciences).
  • In Revenue, your fundraising target is around 230,000 CHF. How many external donors does that include, and how likely is WMCH to secure these funds?
    • We received exactly CHF 167,186.9 from 2015 individuals in 2014. As per the Fundraising agreement our target for the WMF should increase by around 20% this year (tbc), and we expect our income to grow somewhat accordingly (we’re rolling out a new website anytime soon with a simplified payment procedure), so that’s 190-200kCHF coming in already. Not a given, but a reasonable expectation.
    Our French Community liaison will be starting fundraising activities and we *hope* that she will succeed in securing an extra 30 to 50 kCHF. This is a rather low shot because 1) we lack experience in the field 2) it clearly isn’t a full-time occupation (0.4 FTE) and 3) quite a few donors here budget their donations on one year for the next (i.e. 2017). We’re optimistic that she should reach this target, but guarded because we lack visibility/hindsight on the matter.

cheers, KLove (WMF) (talk) 22:29, 13 October 2015 (UTC)Reply[reply]

Thanks for your questions, if anything else please do ask. Stephane (WMCH) (talk) 08:54, 16 October 2015 (UTC)Reply[reply]
Thank you Stephane (WMCH) for the answers! Much appreciated! KLove (WMF) (talk) 05:44, 21 October 2015 (UTC)Reply[reply]

Questions from FDC members[edit]

Questions from Risker[edit]

  • - have you discussed with WMDE/Wikidata about this potential partnership? They are seeking to bring on staff to establish such partnerships. Has there been any work to take advantage of this potential synergy?
    • We got in touch with WikiData this year and invited Lydia Pintscher to the annual conference in July 2015. Due to many other commitments, Lydia was unfortunately not able to attend, but Pavel Richter (former ED of WMDE) held a keynote at the conference and presented WikiData to the Open Data Community in Switzerland. And just at the beginning of this week, we reached out again to Lydia in order to involve WikiData more centrally in next year's conference (sounds to us that Pavel was rather convincing to them) and to discuss potential future partnerships. Long story short: no concrete plans yet with WMDE (it's also up to' management to decide whether they want to pursue anything), but we're already building bridges.
  • Technical tool development, in particular tools related to GLAM. Is there the internal expertise to create these tools? Does it require chapter staff or volunteers? How transferable would these tools be to multiple languages? Would they be submitted as part of the overall MediaWiki/Wikimedia toolset? Some other APG proposals have been made that refer to development of tools for [local] community use, and the WMF has also created a new Community Tech group; how will you work with others to ensure that effort is not duplicated?
    • First thing first: we want to make clear that at this stage we are not (yet) considering developing GLAM wikitoolset II, but rather GLAM-oriented tools. We do already have the necessary internal expertise (we have a staff with 20 years experience in IT, 10 of which in IT governance, + ITIL and COBIT certification) and plan to increase it gradually, but since we do not have the manpower to develop internally, we will be looking for partnerships so as to outsource actual coding work and find synergies with existing projects when possible. At the end of the day, we want to have a blend of volunteers (to whom we're also providing trainings, particularly re:Wikidata use) + staff + partnerships that allows the projects to move forward no matter what. We're not very territorial about where things are made, we just want to make sure that identified gaps are filled and end users (our GLAM partners) are given the solutions they need to justify / pursue / increase their level of collaboration with the movement (no matter where).
    The idea, therefore is to develop IT-supported services that help up work GLAMs - leveraging our network to identify actual needs, figure out what resources are already there, improve on them if need be, and grow the relationship from there. To put it differently, with a very dynamic local GLAM environment, we'd rather invest in nimble tool-based solutions than in a Single Big Tool (i.e. ITIL service design and strategy rather than pure development).
    The collaboration with ATED and with other external tech communities belongs to that long term vision: the Swiss Confederation is pushing a lot for increased collaboration between Universities and external partners (and is financially supporting it) because the thinking here goes that the resulting exchange of knowledge will create the added value necessary to compensate for our high costs. We've developed a good relationship with SUPSI (University of Applied Sciences in Italian Switzerland), and the idea for them is to leverage this and our GLAM partnerships to get direct support from the Swiss Confederation (which they'll repay us through professional tech expertise).
  • UNITAS/Frontenders - this proposal seems to dovetail with the text-to-speech project that is proposed by WMSE, except focused on Wikisource. There are already a range of voice-to-text software products available both commercially and freely, a range of which are currently used by visually impaired Wikimedians. Have you explored whether there are some specific issues with text written using MediaWiki software that requires something further?
    As indicated to Katy above, the feedback we got was that developing Yet Another Tool was not what was needed from the visually impaired: rather, we'll help them work on Wikisource content so as to adapt it to best practices and existing standards. But then, that our own interpretation from Italian speakers: Norway Sweden (and Norwegian Swedish) is a much smaller market and the local ecosystem may not be as developed. In any case, our understanding of the issue is that it's not so much about the mediawiki software than how some of its functionalities are sometimes used in a way that reduces access: that's where we believe that bringing users from this particular sub-group will prove most beneficial.

Thanks very much for your submission. Risker (talk) 02:31, 14 October 2015 (UTC)Reply[reply]

Thanks for your questions! Stephane (WMCH) (talk) 15:56, 16 October 2015 (UTC)Reply[reply]

Questions from Itzik[edit]

  • Global metrics - I understand that not everyone feels like bit-metrics are the right way to evaluate our work (I'm one of them) - but while not using them, we may need to uses or offer others way to evaluate our work - especially for big chapters like WMCH. It's different value when we adding a picture to an article, editing a sentence or writing 1000 words new article. We can just give a goal as numbers of articles affected, but also need to compare the different affect and see the value\costs of each change.
    Well, I agree with what you wrote... but what's your question?
  • Microgrants - you mentioned on the last report that during Q1-Q2 you founded 5 microgrants. On 2016 you planning to fund 20. What the changes you are planning to do in order to almost double them?
    We're currently standing at 10-12 without much publicity - we'll probably get 20 by year's end as we're planning to run a banner reminding people we do offer this service. Essentially it's all about ramping up communication, but it's also about helping established contributors run their projects: we won't shove it down their throats if they don't want it: 20 corresponds to our estimate of the current demand.
  • Swiss - you mentioned under "Wiki loves Y" that it's hard to get to the "Swiss" as Swiss it not a language and your community is spread on a different Wikis - Are you working with Geo-targeted sitenotice? I think it's an easy way, regardless the language to reach your community in Switzerland, no? Or that I didn't really understood the problem...
    Geotargeted sitenotices: yes, but if they allow us to reach Switzerland-based users, they do not allow us to know who they are. Wiki loves Y campaigns really helps us know who is where, doing what.
    Hi, @Stephane! Last minute side note: 3,000 uploaded images - don't you think it is a real low ball? A record single contestants in WMPL's photo competition uploaded more. :D Switzerland is a beautiful country with a huge e.g. Wiki Loves Earth possibilities and your goals could be much more bold. aegis maelstrom δ 18:18, 13 November 2015 (UTC)Reply[reply]
    Hi @aegis maelstrom. The number is quite high actually: I don't know where you got yours, but according to this page, there were only 1735 uploads for WLE 2015 in Switzerland this year, and barely 2300 in France (a country that is 13 times larger and with a population 8 times bigger as ours). Austria got 3600 uploads, so the bottom line is to try to both improve our figure and remain mid-range. Stephane (WMCH) (talk) 10:58, 16 November 2015 (UTC)Reply[reply]
  • Kiwix - While I understand the problem with the current metrics - again, some metrics need to be offered. Maybe the number of people who upgrade? or downloaded more than one time? maybe a survey to get feedback? numbers of volunteers taking part of this project (you mentioned also that 3 microgrants on Q1-Q2 were related to Kiwix.. (I'm just offering - I'm honestly don't really familiar with the project.
    That is something we are really struggling with. Something we could do in terms of survey would be to use the current satisfaction rate for the apps (4.3 on the Android Store, vs. 4.4 for the Foundation's own app: not bad!) - but the problem is that Kiwix is still a volunteer-driven project and that we only support them and cannot directly work on software results (we provide the means to reach certain upgrades, like for Raspberry pi). What would help could be for us to know what the metrics for the Foundation's App are and see what we can take from them (if anything). As for volunteer work, we could also look into increasing the core number of people participating in bug-fixing and development. Honestly though, Kiwix is pretty established yet we still only have a very limited, specialized crowd to work with (3-5 regular contributors for the past few years - remember that for something as big as OpenSSL, which runs 2/3 of the internet, most of the work came from 2 guys only). Our assessment from last year therefore still holds true: we need to professionalize development (hence the fundraising objective).
  • Wikimania - There are already big amount of money allocated for this event from the WMF. How WMCH is related? Are WMCH is the chapter supported the event or WMIT? (asking, I don't know). I didn't checked with Italy - but we need to see that there are not too many organizations and groups funding and supporting Wikimania. Having all your staff in Italy supporting event that is not the chapter event sound really different from all the others Wikimania's we had until now. --Itzike (talk) 13:36, 17 October 2015 (UTC)Reply[reply]
    We made a commitment to support Wikimania with manpower - the event itself has its own budget for everything else- and the organizers have asked us to be there (Esino Lario is in Italy but on the Swiss-Italian border: that's why two chapters are on it). If discussions move forward and they conclude that they'd simply be happy with us doing back-office work, great. But at this stage that's not what we've been asked and we need to budget this. Stephane (WMCH) (talk) 18:40, 18 October 2015 (UTC)Reply[reply]

Questions from Liam[edit]

  • Your current (and proposed) operating reserve is CHF 462,000. This is in effect three-quarters of your expected total revenue. I'm not suggesting that this is 'good' or 'bad', but could you elaborate on this number? Why do you have this reserve (what is it being reserved for?), and why is it the same number for this year and last year even though the total budget size of the organisation has changed? Wittylama (talk) 12:40, 22 October 2015 (UTC)Reply[reply]
    Common (or I should say *Recommended") practice in Switzerland for NGOs is to have between 6 month and a year's worth of cashflow in reserves, so we're actually right smack in the middle of that range (I think the Foundation used to have a similar objective; it's a bit more aggressive now). The number hasn't changed from last year because of this: we don't see a need to take from our reserves, and do not expect to add to them either (which we would have had to had our budget increased substantially).
  • You say that "If we get the same amount of funding as last year it means that we'll have to cut our costs by 10%" but, if I understand correctly according to this table, you have requested an amount that is already about 10% lower than what you received last year. Can you clarify? Wittylama (talk) 14:07, 22 October 2015 (UTC)Reply[reply]
    That was a description of our environment: with us changing nothing, the APG Committee would less bang for its buck simply because of adverse exchange rates. But we've made significant efforts to bring down our costs and improve external funding sources and voilà: we've now more than offset the exchange rate issue.
  • In the "content" program area, you say that you will "Expand the [wikivillage] concept to four languages (als, de, fr, it). 80 participants, 60 of which are new users." However, Wikivillages are listed in the "GLAM: not so good" section of your 2014-15 progress report, where you say that "...the concept is promising, but we think we probably addressed the wrong side of the spectrum...We have therefore decided to limit the planned roll-out to the rest of Western Switzerland and run another pilot project in the Canton of Vaud." These two statements (one about expanding the project in terms of languages, one about limiting the project in terms of geography) seem a bit contradictory. Can you explain? Wittylama (talk) 14:15, 22 October 2015 (UTC)Reply[reply]
    It's a matter of opportunity (and geography): the main issue we had with the first edition is that 1. it is very staff-intensive and 2. kids may not be the best audience for such events. The other thing that happened since the Progress report is that we've been approached by villages a bit all over the place who wanted to partake with their local associations (i.e. the "other end of the spectrum" : mostly older people, but not so old that we need to teach them how to use a computer). Whereas we had a model last year that had us run around schools trying to convince them to include wikivillages in their curriculum and then train the kids to participate, we have a much more fluid one were we are contacted by volunteer associations (after they heard about us from the press coverage of the first edition or through a rather light email campaign) whom we provide training sessions to only if need be. Stephane (WMCH) (talk) 09:15, 27 October 2015 (UTC) Reply[reply]