I'm actually very interested in this as well. Eekim and I have lived and breathed this data for a couple of months, and sometimes being too close to the data gets you lost in the weeds. Given that there are limited resources, is there a reason to prioritize this in another way? What would that way be?

For the record, there's a TON of information in Wikimedia-pedia at China and Africa. I had a very emotional reaction to choosing not to prioritize these areas, because I'm a true believer in the cause and the mission - but the truth is, for me, the Foundation would be spending donor money. When it's given to WMF, it's in the trust that we will make logical decisions that advance the mission in a way that provides the highest return on investment while staying true to our goals and mission. Investing in China and Africa are, for me, not safe gambles right now, and I don't feel like (if I were making the decisions, which I'm not) I could feel good about gambling with other peoples' money in those areas. We don't have the cultural competency or the institutional/governmental support to back it up. I want to extend there as badly as anyone - but with limited resources, it's not something I'm comfortable recommending.

But with that said, I'd love to hear other opinions - but they should keep in mind that the Foundation doesn't have unlimited resources. In my opnion, if there's a strong desire to go in to China or Africa, we probably should figure out how to do it using another entity (chapters, some other part of the movement).

~Philippe (WMF)21:06, 20 January 2010

I think that those areas should be high priority. When I donate money in the fundraiser, I do so very much with those areas in mind. Actually I could have bought a yearly account on an online commercial encyclopedia for the same money each year. But when I decide to use Wikipedia and donate money to Wikimedia instead, I do so because I know that my money can enable someone that doesn't afford to pay for a commercial encyclopedia to use Wikipedia free of charge. I think this should be considered too. Maybe a lot of donors actually reasons the same way and actually wants their money to be invested in less developed regions. Their might be a good idea to make a survey on the motivation behind the donations.

At the same time I understand that investing in those areas not allways will have a very high effect. Some investments that would have high impact in for example Europe might not make a difference at all in Africa or China because the underlying infrastructure and culture that leads to succes in Europe not is present there. I do however think that every low cost-high impact solutions should have an essential priority in the strategy. I think such strategies are what we have aimed for in the Local Language projects Task Force.

Dafer4521:47, 20 January 2010
 

One issue here is that we should be simultaneously developing a five-year plan for the movement and mission, as well as one specifically for the Foundation, to put the latter in context.

Another is that I haven't seen an explicit assessment of the cost & impact stats of various solutions or ideas. From my perspective, if we can define a major priority, whatever the cost, we can find people, networks, organizations, governments, and global superstars to make it happen. And 'low cost/high impact' proposals are always interesting, even when not targeting a core priority.

If expansion in China, India, and Africa are important (simply on the basis of reaching all of manking, as well as current access to knowledge in those areas) but we don't know how to do it effectively, a meta-priority to find better ways to do that [rather than "investing" in a specific type of solution that may or may not work] would be appropriate.

24.91.152.13518:50, 21 January 2010

What guided me in the research I did for the Local Language Project Task Force was precisly cost and impact. However, I found that it was very hard to find numbers that meassured the actual impact a strategy would have. For example the cost for localizing the MediaWiki software was estimated to be $0 (well maybe not zero realy), $90 000 or $900 000 depending on what method that was used to get that work done. But then, how do we know what impact that realy has on the local projects? I asked Siebrand at translatewiki.net for data on the correlation between past localization and project growth, but such statistics didn't exist. Hopefuly it will be available in the near future though, as Siebrand told me that he had spoken with Erik Zachte about this in the past, and that work on trying to make this happen was sheduled to take place in January.

Another recommendation where it is obvious that the strategy will cost a lot of money is the fourth recommendation about having more local servers to rise access speeds. Here it is possible to make the estimates about how much faster the loading times will be. You can have a gut feeling that if average article access time is lowered from 30+ seconds to 10 seconds or below you probably will have larger use of the projects, and thereby also contribution to the projects. But how can you be sure that this realy is so? Maybe other factors play a role as well that won't make the projects grow even if people have instant access to the material.

So even if the price and some parameter that is likely to be correlated with project growth is quite easy to estimate. It is still very difficult to estimate the actual impact that the different strategies will have, because you don't know how strong the correlation between the actual impact and the "likely correlated parameter" is. I would therefore make clear that I am very unable to make any judgement about how large the different recommendations in the Local Language Project Recommendations actually will have, and my hope is that many with a greater understanding of the complexity of the problem can make such.

Dafer4519:35, 21 January 2010

You're right; it's hard to be exact. However, we can make good guesses. Take your example of localisation. We know that localizing our software affects projects across the board, whereas putting a person on the ground in a specific country most likely only affects that country. Now, maybe putting that person on the ground has a greater effect in that country than localizing the software. So the analysis does start getting more complex.

I want to reiterate something. This letter is largely focused on Foundation investments. It makes sense, given the Foundation's limited resources, to focus on things that will have broad impacts. However, because we are a distributed movement, it makes sense for people on the ground to engage in local strategies. One of our big next steps is to flesh out the movement roles questions. Specifically, who in this movement should be doing what?

Regarding cost/impact analysis: The regional analysis largely focused on impact and less on cost (other than China). There's another layer of analysis that still needs to be done. Help us flesh this out. A good starting point is Wikimedia market analysis.

Eekim21:52, 22 January 2010

Yes, I agree with both you and the letter that on-the-ground precense in Africa ain't something to strive for at the moment. One of the first lessons Gerard tought me when I arrived here was that general strategies that will have an impact on a large amount of projects are something to strive for. And I believe that such strategies not are pushed aside in this letter with point 4,5 and 6 under "Synthesis of the issues to tackle with the strategy" and point 3 under "Recommended priorities for the Foundation" supporting this. At least I believe these statements are sufficient for South America, Africa, Middle East and South Asia. How well they address the issues of reaching China and India I don't know though.

What analysis is supposed to be collected? Under market analysis I find a lot of tables that allready are filled in. But is there some specific collection or analysis of data that needs to be done?

Dafer4523:14, 22 January 2010
 

When we put a person on the ground in a particular country, he or she CAN localise for the language he knows. When this person is not part of that countries language(s) and culture(s) he is likely not to be as effective as could be.

When we put someone on the ground in a multi-lingual country, then it should be one of his/her objectives to remove all the roadblocks that exist.. Finding people to localise our software is one of the more obvious ones.

The notion that these approaches are in competition is not that wise; as we understand the processes that promote editors to edit and readers to read, it is only natural to have a person on the ground make those the priorities. Thanks,

GerardM13:43, 26 April 2010

Fair point, Gerard. Many software companies put people on the ground to understand these obstacles, as you point out. This is something that should be included as part of the boots-on-the-ground experiment.

Eekim18:27, 26 April 2010