A Case Media Analysis of Social Networks

Brendan is a researcher for an online business website. He is curious to see how people are using a popular business promotions metric created by Fred Reichheld, Bane & Company, and Satmetrix, called Net Promoter. This tool aims to determine the loyalty of a company’s customer relationships and provides an alternative to older methods of customer satisfaction research. It uses a simple question, “How likely is it that you would recommend our company/product/service to a friend or colleague?” The answers, given on a scale from zero to ten, are quantified with the firm receiving a Net Promoter Score.

This score determines the overall brand loyalty between the firm and the consumer and reveals the likelihood of the consumer recommending the company/product/service.

To determine how people were promoting Net Promoter, Brendan used Twitter, one of the most popular social communication sites in the world. Twitter allows firms to manage their own accounts, giving them the ability to actively communicate with consumers, post updates on products and promotions, and respond to consumer complaints or concerns. It’s the perfect platform for a company to post a link to their Net Promoter survey and assess their customer’s brand loyalty. That data would reveal which consumers are promoters (loyal and likely to recommend the brand) or defectors (likely to go to another brand, not recommend). Brendan conducted a search on Twitter for Net Promoter and downloaded the resulting messages from the previous week. After trimming down irrelevant data, he was left with three categories of messages: tweets, mentions, and replies. This led to the creation of two worksheets, one with data on the messages and the second with data on the users involved in those specific messages.

Next, Brendan conducted a social media analysis by mapping the data to display any observable relationships between users, both directed and undirected. For this, he imported his collected data into NodeXL through his Twitter account and subsequently received a visual map of the social network of those discussing Net Promoter. The map consisted of vertices (a point that represents a specific Twitter user) and edges (visual connections between the users, denoting mentions, replies about, or tweets about Net Promoter). It also highlighted each Twitter user’s degree (the number of users a person is connected to).

I believe the key distinction Brendan needs to make from his map is whether the overall network is directed or undirected. If the overall data set suggests an undirected network, he could posit that the Twitter dialogue about Net Promoter was predominantly conversational, meaning that it actively engaged users. If the data leans towards the overall network being directed, it suggests the relationships between the users are not symmetrical and the dialogue was likely broadcast messages that did not necessarily receive replies. To identify key vertices in his map, he should rank each vertex’s degree to understand the full extent of the user’s audience. With high degree vertices, the question remains: is there a dialogue with feedback or are there simply messages with no conversation attached? To answer this, Brendan should first rank the degree of each vertex, then ascertain whether the vertex created a directed or undirected network.

Did you like this example?

360

15