User talk:Melamrawy (WMF)/Community collaboration best practices

Questions: User:Halfak (WMF)

Hey! So I set aside some time to answer questions this morning, so here I go. One quick note: I don't really have a funded project team. Most of my team is either volunteering or I have secured them funding by working with IEG or external entities like WikiEd and Wikimedia Deutschland. So, we're a very non-traditional "Product Team" and I have a to fill a lot of roles as people come and go.

1. In your platform, what types of different user groups exists?

Hmm... I'm not sure what you mean by "platform". I'd say that my primary "user base" consists of:

• Tool developers -- Wikipedians who use python, MySQL and Javascript to build automated systems that help Wikipedia editors get their work done.
• Patrollers -- Wikipedians (who often use automated tooling) to review and reject bad new edits and new page creations. "Bad" usually means vandalism or spam.
• Mentors/socializers -- Wikipedians who seek our good-faith newcomers to help and support. They are often involved in mentoring initiatives like the en:WP:Teahouse.
• Newcomers to Wikimedia projects -- These users are unusual since, by the time they find us, they aren't newcomers any more. However, a lot of my work centers around their concerns.
• Researchers -- Covering fields from Literary Critique to Computer Science/AI, these researchers often come to Wiki(p|m)edia for different academic reasons, but they often end up asking similar questions and they need access to similar datasets. Individuals in this group have a mixture of backgrounds. Some of them are trained academics. Some are amateurs who just started running studies. --Halfak (WMF) (talk) 16:49, 12 December 2015 (UTC)

2. How far do core community members affect product decisions?

As much as I can get "community members" to collaborate with me. My usual pattern of responding to bug reports and feature requests is to encourage the reporter to either fix it themselves or write a proposal for how it should be fixed. When I work on a new system or fix a problems, I document my activities aggressively so that anyone can see what I was thinking and the results of my work. My own prioritization is substantially driven by what people ask for (when people are asking) and what I think is best (when people are not asking). Generally, I find that the major limitation in "community members" affecting my decisions is the time and energy they have to devote to formalizing their concerns and ideas into actionable proposals. Our volunteers are very busy people. --Halfak (WMF) (talk) 16:49, 12 December 2015 (UTC)

3. How far do you adapt your products based on community suggestions?

Hmm... measuring this distance is difficult, but I can imagine some things. So, a request to substantially increase the complexity of a project is usually met with a discussion of the limitations on the resources we have (All work and no play makes Aaron less productive). However, if someone wants us to look into a different direction or try something in a slightly different way, I usually go out of my way to take the first step (make a phab card/mock/dataset/whatever) and see if the requester reciprocates by carrying it forward. IME, this is a really good way to bring new people into the project. By brining volunteers into the project like this, I hope to have them drive the directions that projects take. It's hard to say which direction projects would have taken without following this pattern of pulling in volunteers, but I would guess that it's substantial. I always make it a point to invite anyone who wants to work to weekly meetings and to encourage them to take charge of what interests them. --Halfak (WMF) (talk) 16:49, 12 December 2015 (UTC)

4. As a casual observation, how happy is your core community with the product decisions you make?

I'd say those who notice are very happy. We often receive drive-by negative feedback, but that often leads to productive contribution along the lines of the strategies I described above. The main problem that my community has with the work that my teams do is that it isn't getting done fast enough. That's a good place to be an a volunteer-rich environment. --Halfak (WMF) (talk) 16:49, 12 December 2015 (UTC)

5. How do you release new features, that didn't stem from community suggestions?

Our whole process is open so "didn't stem from community suggestions" is a bit blurry. I'll interpret that as, "anyone who doesn't look at our phab board or progress reports doesn't know it's coming". If the feature is backwards compatible, we deploy it with an announcement and make time/space to address any immediate feedback. If the feature is going to break some access pattern or assumption then we organize a public discussion and CANVASS for participation. So far, it hasn't been necessary to make a formal RFC, but I imagine that may happen at some point. Sometimes we break something in a release. When that happens, we write a report about what happened and why. Then we leave status updates as we look into the problem and solve it. --Halfak (WMF) (talk) 16:49, 12 December 2015 (UTC)

6. How do designers on your team, (or you, if you are a designer) perceive community comments?

I wouldn't say that we have professional designers on our team, but we're all designing things anyway, so I guess that makes us amateur designers. Generally, when there is a complaint about design, I try to figure out if it is an aesthetic preference (still valid) or a missing affordance in the software. If it is a missing affordance, we'll fix it ASAP. If it's an aesthetic preference, I'll redirect by asking for a proposal on how to make a change. Then, we'll iterate openly on the design proposal. This is partially to redirect the complaint constructively and partially to try to have a make sure we have an open process around design changes. Recently, we've had MGalloway volunteer to help us with some design work. I'll probably have a better sense for how to answer this question after we work together for a while. :) --Halfak (WMF) (talk) 16:49, 12 December 2015 (UTC)

7. How would you wish for things to change?

I'd like to have some more flexibility for engineers and designers at the Wikimedia Foundation so that when we get some big asks with big impact from our communities, I don't need to convince engineers/designers to work after-hours to help us. I'm very happy to have the problem of convincing people that the work is worth their time is left up to me. I think that such a pattern would match the work patterns of our volunteer community much more closely and would allow for us to be more adaptive to their needs -- and to ideas that originate from outside the WMF. --Halfak (WMF) (talk) 16:49, 12 December 2015 (UTC)

8. Any comments you would like to add?

I'm thinking totally selfishly here. I want to see my projects succeed. I need help and our volunteers are a massive resource in this regard. The best way I have figured out to take advantage of these volunteers' expertise and energy is to share ownership. Got a complaint! Cool. Write a proposal and let's start talking about what *we* need to do to get this resolved. I don't view my projects like a Product or Service provided to a User. Instead, they are open projects that blur the line between the user and the contributor. --Halfak (WMF) (talk) 16:49, 12 December 2015 (UTC)