« Event review: Twitter’s Impact on Media & Journalism | Main | The motivations of influencers and amplifiers: how content becomes prominent »
Uncovering high-value applications of organizational network analysis
New Scientist has published an interesting article titled Email patterns can predict impending doom, which reviews findings by researchers at Florida Institute of Technology. They, as many researchers, used the email logs from Enron, which have been made available for analysis by federal investigators.
The key finding from the research was that the number of active email cliques, in which groups exchanged emails between each other but not outside, went from 100 to 800 a month before the collapse of the company. This appeared to reflect decreasing trust across the broader organization and increasing stress. This indicates that very strong indicators of organizational health can be gleaned from network analysis.

Network analysis by Advanced Human Technologies of top executives in global corporation
As indicated in the article, one of the key challenges of organizational network analysis is obtaining email logs, even if the data is disguised. Advanced Human Technologies has been applying social network analysis to enhancing organizational performance for most of the last decade, using both surveys and digital trails to provide management insights and identify high-value initiatives.
There is now a very solid body of research on how to apply insights from network analysis to improving organizational performance. However in the big picture this is just the beginning. There are very rich insights - and rewards to those that apply them - in understanding organizational dynamics. We're keen to find organizations that are willing to share encrypted data in order to generate unique insights. Full privacy is maintained, while potentially creating significant competitive advantage.
For the most current insights and trends in the living networks, follow @rossdawson on Twitter!





















Have they actually used this technology to predict something *before* it happened. Confirming (afterwards) that there were signs available before an event does not neccessarily a useful prediction system make.
I must admit I haven't read the original paper but even if we ignore the hurdle of privacy issues I suspect that analysing network traffic such as emails is going to be a rather blunt instrument. It might very well confirm that something is happening but the value lies in knowing *what* it is.
/Johan
Fair comment Johan - there is a long way to go to get the sorts of correlations and deep insights that will create value. The far greater value comes in text analysis, not just emails volume and patterns. Interestingly, Tacit Software, a major player in this space, was acquired by Oracle six months ago.
Your article is nice.Thanks for your sharing,it helps me more.I will look forward to your more wonderfull articles.Have a good time.