Grants talk:Project/Commons app/Commons app Android v4/Midpoint

Ready for review

@Mjohnson (WMF): The midpoint report is now ready for review. Thanks! :) Misaochan (talk) 11:39, 8 December 2021 (UTC)

Follow-up questions

Hello @Misaochan, it was really nice to connect via video call. Thank you for sharing your report updates and learnings. It was fascinating to read about you/your team's work and intent to build an app that can be widely used by the community. Your dedication in consulting and taking in feedback from the community and WLM international organisers is highly valuable and comes across in your report.

We are curious about the following and was wondering if you could shed further light on:

- Would you be able to help us understand the geographical distribution of people who downloaded the app? as well as users (i.e. upload photos) of the app? What observations/trends can you draw from the data?

- When consulting and taking feedback from stakeholders on adaptations for the App, what's your thought process in deciding which fixes to work on (and how much resources to spend on them) and how do you prioritise them (i.e. decision making process)?

- Under the FAQ segment of the App (we tried downloading it on our mobile =), there's a question and related answer on

"Despite having entered the map where I'm interested in, I can't find the monument I'm looking for and even it does not appear in the list when searching by name. How is this possible?"

To have a monument participating in Wiki Loves Monuments Its required that it's owner or its legal custodian authorises the participation. If you know, or if you are, the owner of the legal custodian and you wish the monument participating in the contest, please refer to instruction available in this page or contact XXX"

How do we better sync existing collection of monument photos already available on wikiprojects with the App so that contributors can better focus on filling in the content gaps to maximise effectiveness?

Thank you and looking forward to hearing from you.

JChen (WMF) (talk) 11:31, 17 February 2022 (UTC)

Response to questions

Hi JChen (WMF), it was lovely to chat with you too, and thank you for your kind words!

Would you be able to help us understand the geographical distribution of people who downloaded the app? as well as users (i.e. upload photos) of the app? What observations/trends can you draw from the data?

I have collected statistics of active installs by country from the Google Play developer console today, parsed them, and displayed the top 20 countries in a pie chart and a bar chart. The total number of active installs is 9,069.

It may be worth noting right off the bat that we have installs in 148 countries and therefore the data has a very long tail; however, there is no effective way to display all of the countries, and it would be difficult to analyze, so we have just chosen the top 20 to display and discuss.

Looking at the data above, it seems there are several factors at play. On the one hand, the overwhelming majority of Wikipedia (and by extension, Commons) contributors are from the global north, therefore we would expect there to be an overrepresentation of these countries in any app or software aimed at contributing to Wikipedia/Commons. So we do have several global north countries in our top 20, but the global south countries are not far behind, and certainly much more prevalent than expected. For instance, the wikipedia link above says "The only country not in Europe or North America in the top 10 is India (3%)"; however, the demographics of our users are very different. We have 4 global south countries in our top 10: India at 2nd place (with 890 active installs - nearly 10% of our total), Brazil at 8th place (185 active installs), Indonesia at 9th place (167 active installs), tied with Nigeria.

My hypothesis is that this is due to the fact that we are targeting the Android platform. This is a double-edged sword - the large variety of Android devices makes debugging much more difficult, but the main benefit of Android is that almost every phone that isn't made by Apple uses the Android OS. This means that most of the low-cost phones on the market use Android, and most mobile phone users in the global south are likely to use Android phones. I believe this is the reason for the relatively large % of global south countries among our users, compared to the norm.

Re: uploaded photos, we are not able to obtain direct statistics for every upload, as many Commons uploads are not geotagged (the users can remove geotags for their own privacy, and at any rate we do not have a script to segregate uploads that do have geotags). We are, however, able to generate a map of uploads made through our app to Wikidata items lacking photos that have a geolocation, as shown below. This map follows the same trend as active installs: there is a relatively high representation of global south countries, although of course it can still be improved.

When consulting and taking feedback from stakeholders on adaptations for the App, what's your thought process in deciding which fixes to work on (and how much resources to spend on them) and how do you prioritise them (i.e. decision making process)?

We collect feedback from users and Wikimedians from various entry points: mailing lists, reviews, emails from users, some users will create issues on GitHub themselves - but the end result is that they all end up in the central repository as GitHub issues. We have a group of advisors, volunteers and core developers who have been with the app for a long time in various capacities (testing, development, repository maintenance, power users, prominent members of the Wikimedia community) who will flag issues that they perceive to be of high importance. If a volunteer developer chooses to work on any issue for free, we will almost always allow it as long as it benefits the app, so there is much less prioritization that happens there.

For work that is actually funded, the core developer team and advisors will discuss the direction that we wish to take with the app prior to the drafting of each grant proposal (grants are our sole source of funding). Issues will usually be prioritized based on the strategic direction that we wish to take - i.e. targeting global south users, benefiting monuments, etc. After the final draft is agreed upon by all core developers and advisors, it is posted here. Feedback is solicited on the proposed improvements from the Wikimedia community via mailing lists and on-wiki discussions, and also from the grants committee. Often, as with this particular grant, WMF technical staff will also weigh in with their own requests for the app. Edits will be made to the proposal until it is satisfactory to the majority of the Wikimedia community, grants committee, and WMF, and when the proposal is approved, the priorities are essentially locked in for the remainder of the year.

We do also fix major issues that crop up that aren't stated in our grant proposal, and this is done on an ad hoc basis. Prioritization for this basically depends on how urgent the request is, and how much time we have to spare from the priorities that were locked in with the grant proposal.

Under the FAQ segment of the App, there's a question and related answer on X

As discussed via email, that FAQ section is from the WLM Italia app, which isn't maintained by us. Our app can be downloaded from the Play Store here, and we have our own FAQ.

But to try and answer the question: We feel that with regards to syncing of monuments already available on wikiprojects with our app, the best thing that can happen for our app is for the WLM project to have a universal system of identifying and categorizing monuments, as well as standardized WLM templates. That would allow us to enhance our WLM feature to be truly usable by every country participating in WLM. However, we understand that this may be difficult for various reasons which are best discussed with the WLM international organizers.

Please let me know if you have any further questions, and thanks for the review!

Best regards, Misaochan (talk) 15:41, 17 February 2022 (UTC)

Mid-point report accepted

Hello @Misaochan, thank you for your quick response and also for sharing your thoughts and analysis. We have reviewed and are accepting your midpoint report. Thank you and your team for your efforts and for being so invested in the project. We (as a movement) are all the better for it. =) JChen (WMF) (talk) 09:11, 18 February 2022 (UTC)