The Tuesday afternoon session on Organizational Issues attracted me for the first session on knowledge management and help desks and kept me for the discussion of how organizations need to think and change to deal with Computer Supported Collaborative Work.
Behind the Help Desk: Evolution of a Knowledge Management System in a Large Organization by Christine Halverson and Tom Erickson of IBM; and Mark Ackerman of the U of Michigan.
The Diffusion of ReachOut: Analysis and Framework for the Successful Diffusion of Collaboration Technologies a research note from Vladimir Soroka, Michal Jacovi of IBM Research in Haifa.
Return on Investment and Organizational Adoptions a research note by Jonathan Grudin of Microsoft Research.
Leveraging Social Networks for Information Sharing a research note by Jeremy Goecks, Elizabeth Mynatt of Georgia Tech. Describing their Saori project.
The Behind the Help Desk work described a two-year evolution of an internal help desk / knowledge management system, centered around an FAQ application that was inspired by various organizational studies. In this particular world, the internal help desk had turnaround times on questions of weeks, due to the complex nature of the requests both from a technical side and the organizational side. For example, some suggestions were technically feasible but would create massive havoc in the client organization if implemented. As a results, the help was much more than simply "do this thing" and included advice around how to manage a particular group interaction. The FAQ system grew out of the internal experts' recognition that there were frequently common core aspects to these otherwise unique problems, and if they could systematically record and publish this information it might provide benefit to the entire group.
The valuable observation at this point in the project was that it has changed and grown and diverted from its original goal. The presenters claimed that rather than chaos, that this is bricolage: a natural process of "putting tools together in new ways to accomplish new needs and make unforeseen new uses possible."
The diffusion of ReachOut presentation discussed the diffusion of a collaboration system, ReachOut, in two different environments where it succeeded and didn't. They gave some explanation of why this worked and some thoughts about what this implies for future projects. Those important aspects included
- A clear definition or description of the problem to be solved.
- A link between the problem and the tool. It doesn't work if the tool is just "interesting."
- Relative advantage of the new system
- Compatibility of the change / tool with the environment
Surprisingly, there were some aspects that didn't have an effect in their small sample study.
- Promotion of the tool (change) through many channels. This turns out to be not quite as important.
- Support of management was weak in both.
Return on Investment and Organizational Adoptions presentation was a "this is reality" discussion of how organizations think about ROI measures. Rather than it being a "death of the project" question, it really is a small part of the decision-making process in organizations. Grudin talked about the factors that go into decisions around technologies (performance, efficiency, money, execution, support), suggesting that ROI only addresses one aspect: execution. Essentially, ROI is one piece of evidence that a project or technology will create benefit for the organization.
The Leveraging Social Networks for Information Sharing presentation had the best description of social networks that I've ever heard, unfortunately I didn't write it down. All I remember is that he said in one or two sentences what I have flailed to describe. Something like, "your social network is about who you know and how you know them and how you can get things done through them." They've built Saori to help visualize awareness, mutual visibility, mediation and social accountability. Some observations: There is semi-public information that one wants shared less and less as you walk further out in the network. There is private information where you want to control how and where that gets distributed. Current social network tools don't necessarily appreciate this, though LinkedIn at least has a way to block some types of inquiries (show network; allow requests).