Fred Turling, SHS Viveon AG presented one of the first session of the morning here at the eMetrics Marketing Optimization Summit. Fred's topic was on Social Network Analysis, which starts off the Social Media Metrics track here at the conference in Washington DC.
Social Network Analysis (SNA) can become a pretty deep subject. Looking at the value of participants in social networks such as networks is becoming a pretty hot topic. It's no longer about which community can best benefit from the relationship building in a social network, but honing in on which community members are the most valuable to build the relationships with.
Fred went on to describe that there are three main approaches to analyzing social networks:
- Web Analytics - metrics & analytics of reach, interest, conversion funnel;
- Business Intelligence (community platforms) user generated content, user characters detailing daily usage patterns;
- Social Network Analytic & Data Mining - focusing in on social patterns, etc.
Looking at the background of where Social Networks come from, it's not new at all. The term social network was introduced in 1954 by JA BArnes and refers to a social struture, in which individuals are connected to each other via relationships. That can take on a variety of meanings from written letters, to post it notes and now even to emails and IM's. In the contet of current media networks 4 distinct types can be identified:
- Communication - directly targeted communication (email, im, chat, sms), discussion oriented (message boards, blogs)
- Contact Management (social software - facebook, linked in)
- Interaction - nline games (world of warcraft), commercial or service auctions (ebay)
- Collaboration - collaborative indexing (digg, delicious, flickr), Wikis (Wikipedia)
The hot topic in SNA is the "Small World" Phenomenon, within a social network, people are "not far away" within 5-6 relationship to get to a point.
- For this phenomenon to apply, a network must have a particular structural characteristic, which many real world networks share (social -communities, power grids, www) it must be scale-free
- In simple terms, scale-free networks consist of a few highly connected "hubs" and large number of moderately connected "normal" nodes
- Communication paths often use the"back bone" of hubs and thereby connect any sender to any receiver via only a few intermediate steps
- If a social network has these characteristics viral mechanism can be used to support a marketing-to the few approach
Through their research, Fred found that people interact differently from community type to community type, as well as from community to community. Generally they found that on forums, people tend to have many conversations of a short nature, on collaboration sites such as wikis (Wikipedia for example) the users tend to have few conversations that are much greater in length and nature.
Social Rank also comes into play when looking at who is most valuable in the community. Social network individuals can be ranked according to their status or importance to the community. A straightforward model of social rank may start with the assumption that an individual's status or importance depends on the attention that the individual receives himself and on the status or importance of people in the individual's network neighborhood.
An individual's social rank has implication in forums. Members/customers can be segmented according to their Social Rank. Looking at the analysis, the distributions of social rank may reveal of small segment of top users with extremely high importance of the community (HUBs, Multipliers). This means that the traditional Business Intelligence task of behavior modeling must now take into account not only and individual's own behavior, but how that individual's behavior can influence their network. Social Rank is the first way to segment users in a social community. The more active, the more attention they get. helps to build a priority model to those who are highly engaged.
Lastly, in this highlight of the session (there was a TON of great information I cannot give total justice to), Fred addressed the "distrubution of frustation". This is when customers or audience feel frustrated. These members of the social networks who rank high with Social Ranking, these "hubs & multipliers" can distirubute frustration within a very agressive way. An important issue in social networlks is whether one individual's behavior and characteristics tend to spread to his network neighbors, and how that can affect businesses in the end.