Tuesday, January 22, 2013

Social Network Analytics


When I started to look at this topic, I saw this metaphor for several leading social networks: Twitter is a bar, Facebook is a coffee shop, and LinkedIn is a library1. All illustrated the basic nature of social networks as platforms where people seek information and social engagement2.
Given the amount of time people spent on checking and updating Facebook, Twitter and other social networking sites, it’s obviously a goldmine of useful and interesting data to be analyzed. And from my personal experience, a friend’s recommendation or positive comment is stronger than any repeatedly seen advertisements. So this is definitely becoming new media of increasing brand awareness and perception, and potentially, sales amount.
However, as intuitive as it may seem in terms of the usefulness of the information that could be revealed by doing analytics on social networks, it still remains a question whether it is worth the effort to dig into the extremely complicated data (mostly textual) present on Facebook, Twitter, etc. How can we use that to our best interest? Is there anything that we are missing out when we are talking about social network analytics?
In short, social network analysis could be defined as “the study of the social media metrics that help drive business strategy3”, thus leaving us the big question that which metrics should be chosen as the evaluations for Key Performance Indicators. Before we get to that, we may want to clarify the value of doing social media analytics first. Given the randomness in the data, most of the data collected from social networks may not be relevant to our specific business goals. So it’s probably safe to stay conservative regarding the short to medium term benefits of social media analytics4.  Another problem is that the predominance of Facebook, Twitter and LinkedIn type of social network sites kind of clouded our cognizance of the term “social network”.  We need to recognize that there is another type of social media which could be potentially even more valuable in terms of the ease of doing analysis and getting relevant information, which is online communities. Online communities are different from the social networks based on the types of conversation happening on the platforms, as well as the ways they are structured5 (see Figure 1 & 2).
                                                  
                                              
However, these two types are not mutually exclusive. As we can find in Facebook groups and people following the same public pages, they are just like online communities in regards of the fact they are brought together by common interests. And the fact that one person can be part of different communities would contribute to developing insights and campaigns for lead generation. 
Last but not least, when we are doing social analytics, it’s important to not only concentrate on visible volume and size of interactions happening on that specific platform. Invisible information like browsing behavior would reveal more customer online behavior. I’ll get to more detailed and in-depth metrics and considerations next time.


Reference
22.      Using Web Analytics to Measure the Activity in a Research-Oriented Online Community, Catherine Dwyer, Yi Zhang, Starr Roxanne Hiltz
33.       Actionable Social Analytics, April 2012