Article Review: Determining Influential Users in Internet Social Networks

…as the concept of “business” has developed through social networking—namely, that personal branding is a business entity—being cognizant of these business strategies is important to increase your personal brand in the most efficient manner possible. This can be achieved by directing your efforts to the influential users of your site (or consumers of your content).

Trusov, Michael, Anand V Bodapati, and Randolph E Bucklin. “Determining Influential Users in Internet Social Networks.” Journal of Marketing Research (JMR) 47.4 (2010): 643-658.
http://search.ebscohost.com.ezproxy.mnsu.edu/login.aspx?direct=true&db=ufh&AN=52119022&site=ehost-live

Abstract:

The success of Internet social networking sites depends on the number and activity levels of their user members. Although users typically have numerous connections to other site members (i.e., “friends”), only a fraction of those so-called friends may actually influence a member’s site usage. Because the influence of potentially hundreds of friends needs to be evaluated for each user, inferring precisely who is influential—and, therefore, of managerial interest for advertising targeting and retention efforts—is difficult. The authors develop an approach to determine which users have significant effects on the activities of others using the longitudinal records of members’ log-in activity…The approach identifies the specific users who most influence others’ activity and does so considerably better than simpler alternatives. For the social networking site data, the authors find that, on average, approximately one-fifth of a user’s friends actually influence his or her activity level on the site.

Research questions

  1. How much influence do friends, or friends of friends, have on user’s site activity within a social network?
  2. Does influential site activity correlate to a demonstrable business profit? [monetary or otherwise]

Research methods

This study tracked daily site log-in activity from an anonymous social networking site.

The users were classified into three distinct groups:

  1. Level 1 network: a user is part of a Level 1 network through an established “friendship” link (i.e., an accepted “invitation.”
  2. Level 2 network: a friend that is not part of the user’s Level 1 network but is in the Level 1 network of one of their friends.

For this study, full profile information was observed for Level 1 and Level 2 networks: networking goals, number of friends, and number of profile views.

Log-in activity was taken as a function by the following:

  1. The user’s characteristics.
  2. The user’s past behavior on the site.
  3. The log-in activity of the user’s friends.

Data obtained over a 12-week period

  • 330 users tracked
  • 29,478 total Level 1 friends
  • 2,298,770 Level 2 friends
  • 2.48 average number of log-ins per day

Results

The research method used to examine users according to a user network level is explained by the authors as being sufficient to represent the activity level of Level 1 and Level 2 friends independent of the activity of Level 2 friends. However, “This does not imply that a [Level 2] friend has no effect on a user. Rather, it means that a [Level 2] friend has an effect only through a [Level 1] friend” (p 647). This is important to note because the contrary could imply that having many friends would make a user influential independent of the effects of other users. One way the data showed this was by noticing that regardless of number of friends, if a user’s site usage fluctuated (went substantially up or down) and there was no recorded change in Level 1 or Level activity level, that user was not influential.

The first important result found that a user’s site usage was dependent on the site usage of their Level 1 and Level 2 networks (and in this instance, network is synonymous with friend).

Who influences whom?

One-fifth of a user’s friends actually influence his or her activity level on a site

However, of the Level 1 users that do not have friends that influence them, 32% are not influenced, statistically, by their Level 1 or Level 2 network’s site activity.

The analysis of the Level 1 network’s full profile data and site log-in activity finds these results:

  • Female friends tend to have more influence on male users
  • Users who have been members of the site longer have more influential friends
  • Users are more influenced by member friends of the same ethnicity
  • Users who have stated dating objectives have fewer friends who are influential
  • Friends who are older than a user have less influence

Level 2 users, on the other hand, have significantly less impact on other’s site activity but may still display some level of influence. The importance of this is explained by how some studies use the “simple metrics” of site activity alone to determine influence—whereas this study uses a variable equation (“univariate friend”) in conjunction with simple metrics to determine the influence of top users on a site. Only the top users are important to analyze because even for the top 10%, there is a sharp decline in influence between the highest and lowest.

Why care about influence – Who cares?

The results of this study, regardless of the methodology they propose to use, are useful to purveyors of content within, and between, social networks. “Between” social networks is important to note because this study did not study the impact that users from outside networks have—as we all know, all networks are connected. Users within Facebook are connected to, and are presumably influenced by, their Level 1 and Level 2 networks in Twitter, LinkedIn, and every other social media use or social network in which they participate.

For content creators, targeting the influential users is important. This study helps classify who influences whom and this is helpful for producers to be cognizant of when interacting in and throughout their social networks. For example, if you are presented with two comments and you can only reply to one, taking this study’s metrics into account will help you decide which user has more “influence” and will potentially increase traffic to your site. Additionally, using the model presented can help you to know who has the potential to refer an influential person your way.

Content producers as business entities

This study also helps put into perspective how content purveyors should manage users. For instance, conventional business thinking has established that keeping an existing customer is less expensive than obtaining a new one. The same is true for your content consumers. The potential negative impact that losing an influential user has is not limited to the loss of this single user’s presence or any monetary concerns (e.g., pay per click); no, the site usage or content consumption from all the Level 1 and Level 2 users of the lost influencer will be affected as well. This could be seen by a drop in site views, fewer comments, less “reach,” and ultimately a possible exponential drop in content consumption.

This study focused on the direct business need to target influential users for marketing reasons. However, as the concept of “business” has developed through social networking—namely, that personal branding is a business entity—being cognizant of these business strategies is important to increase your personal brand in the most efficient manner possible. This can be achieved by directing your efforts to the influential users of your site (or consumers of your content).

Advertisements

Leave a comment

Filed under Article Reviews, Social Media

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s