The first thing we need to think about is 'who' is it we want to get the impact analysis to: farmers, extension agents, development professionals. This should provide some context for whom we are trying to target since it is important to think of audience rather than at general levels which can get confusing and lead the dilution of impact of the system. For instance if we look at LAD and Lao44 as to examples of "clearing houses" or repositories. LAD is for a wider audience and has both Lao and English and is focused on both research and development. Lao44 is solely focused on lao language materials though for a broader range of topics. However, the fact that it is in Lao language makes it much more appealing to the Lao people.
 
IN terms of classification and organization of information, i would argue that we need to need less classification and more ways to make it easeir for users to access information. The current systems employed at a global level are too generic and complicated to be used at the local level. Again, with the case of Lao44. There are no standards really but rather a clearing house where information is stored. How it is used is another issue.
 
SOmething that is forgotten in ICT4D is the whole point of socialiazation and face to face interaction. THis is important to become a 'trusted source' of information. There are too many cloud systems, repositories or standard solutions out there but they are not linked to building 'communities'. Again, i think lao44 and LaoFab are instructive. The first step was developing a community through the discussion groups. the repositories came second. I think this is important becuase it is first important to develop trust and then information will be used and shared more effectively. I think the same is true with the sourcebook and other products we made where the partnerships propelled use in the end...