User:Staszyk/Community Polls

Objectives

In March 2007, Gandalf conducted a survey among Mozilla localization teams worldwide, which helped us learn more about Mozilla localizers, the scope of work they do, the process they employ, the problems they encounter and the successes they are proud of. The survey revealed also -- as a side effect -- that our communities worldwide seek out to provide Mozilla with their feedback. They used the survey to report and alarm about the problems, probably hoping tat Mozilla would act upon it.

Now, I've been thinking about two things:

1. how to use the data Gandalf collected in March?
2. how to learn more about the communities and their problems? (But in a way that's superior and more beneficial to both sides than having a report-a-community-problem@mozilla.com e-mail alias.)

Target

Let's establish some vocabulary.

1. users -- all the people who use the Mozilla products
2. community members -- all these fantastic people who try to change the Internet by spreading the word among their friends, providing user-to-user support on the forums, writing on their blogs etc.
3. localizers -- people who translate Mozilla products as well as the Mozilla online content to their language.

In each of these groups there are more and less active people, but it is important to remember that the l10n teams are not groups of elite, most active community members. It may be so in some cases, though.

Gandalf's survey was targeted at the localizers. What I would like to do is to have feedback from the most active community members, as well to reach out for the community.

The concept

1. translated into local languages
   1. can be done like /$locale/l10n/main.lang for example
2. conducted frequently (new poll every 2 weeks?)
3. poll's lifetime = a month (1 week to localize the questions, 2 weeks to colect answers, 1 week to publish the results in English)
4. data collected with a web interface (preferably @mozilla-europe.org)
5. standardized questions (very few question types (3))
6. propose-your-question feedback option
7. short (optimally 1-3 multiple-answer question with max. 10 answers (in random order) to choose from)
8. no open questions

Issues to address

1. repetitiveness -- we should make sure that:
   1. the questions are short, so that the responders aren't discouraged to answer
   2. the questions vary, so that responders don't have a feeling we're asking same things over and over again (which, again, would discourage them)
2. 'awards' -- we should propose something in return for the provided feedback
   1. not T-shirts and such
   2. other people's answers; i.e. publicly available summaries of the surveys
3. getting involved -- by answering, the responders give us their precious feedback, and also help shaping the community. We should emphasize that.
   1. have users-proposed questions asked
   2. featured question wins a t-shirt to its creator
   3. makes the surveys both a Mozilla-communities and community-community communication channel
4. asking different people -- each time, we will be asking a different set of people , under the different circumstances, and at different time, who are only a part of the community
   1. use significance test to generalize the results for the whole community
   2. try to make responders 'loyal'
      1. e-mail newsletter
      2. RSS feed with new questions
      3. blog with the results and discussions about them
      4. litmus-like accounts??? (displaying number of surveys the user took and/or proposed) -- total anonymity essential (however we could benefit from being able to identify the same responder through a series of surveys, without being able to link their answers with their profile)
   3. but: evade the situation in which the only people to take the nth survey are the loyal responders;
   4. embeddable <iframe/> code to put in the blogs, forums, news items?

Examples

• community meetings
• local Mozilla meetings
• time resources
• money? (not sure Staszyk 08:37, 19 September 2007 (PDT))
• means of communication within the community (forums, wikis, irc, irl, blogs/planets)
• command of English
• programming skills
• studies/Employment
• sex, age

Likert scale example questions

1. My friends ask me often to solve their computer problems: [strongly disagree, ..., strongly agree]
2. In the past 1/3/6 months I have visited (or read RSS from): [planet.mozilla.org, blog.mozilla.com, spreadfirefox.com etc etc]
3. My community provides very good user-to-user support
4. I know who to address myself to in the Mozilla Community if ...
5. My community is often asked by media about ...
6. I am proud and satisfied with the localisation of Mozilla products into our language
7. I feel that the community is growing
8. ... that the situation/atmosphere is good
9. As far as on-line content is concerned, for your local market, it is very important to concentrate on: [MDC localization, SUMO localization, Mozilla Blog localization, Spread Firefox localization, extending local knowledge base]