Who can advance your viral marketing campaign the furthest, and how can you find these key individuals on social networks? The answers “can predict the success or failure of social commerce initiatives,” says Morad Benyoucef, an associate professor of management information systems and a contributor to the IBM Centre for Business Analytics and Performance at the Telfer School.
Benyoucef’s ongoing project, Increasing Marketing Campaign Performance: Using Influential Users in Social Networks, has the potential to make a real impact, with particular relevance for companies using predictive analytics to market their products and services. The social media interactions that shape what people buy are a gold mine to companies seeking to make their campaigns more targeted and cost-effective, Benyoucef notes. The evidence for this is all around: Nielsen’s latest Global Trust in Advertising report, for example, found that 92 percent of consumers say they trust recommendations from friends and family above all other forms of advertising.
The notion of influence applies as much to conventional marketing tactics as it does to emerging approaches, such as promotions that provide consumers with opportunities for interaction (e.g., coupon offers where the redemption value increases the more people who share the coupon) and predominantly crowd-sourced marketing, where advertisers tap into social expressions about products or offers which they can then use to communicate a brand’s feel.
Facebook provides a ubiquitous illustration of influence; the integration of marketers and merchandisers on the social network allowing them to target customers based on what they “like” and “share.”
The ability to broadcast to your friends “that you just purchased the latest iPad” is very powerful for businesses, and influence is central to this, Benyoucef says.
“If the two of us are friends and I purchase iPad, chances are I can influence you to purchase an iPad.”
Electronic word of mouth
Part of the research project involves propagation, or how far or how fast users distribute messages in a social network. Amir Afrasiabi, a PhD student in computer science, has investigated how people repost or share content differently based on whether they are friends or followers of a channel.
“Knowing the difference between subscribing to a channel and being friends with a channel is important because that’s how you target people – based on the kind of relationship they have with others,” says Afriasiabi.
Afrasiabi, who recently published a study titled, "Propagation in online social networks" in the Journal of Information Systems Applied Research, is pursuing doctoral work under Dr. Benyoucef’s supervision which focuses on, among other questions, new algorithms for community detection.
Benyoucef says computer scientists and marketing researchers have much to contribute on the question of leveraging influential users. It’s difficult to compare existing commercial platforms in this regard because their algorithms aren’t public. For this reason, a prototype model developed within a university setting can play an important role in filling the knowledge gap. That’s the anticipated contribution of the Telfer School project: the development of a complete model with the potential to be used to locate influential users in a social network.
“Being able to identify influential users in a social network would provide business with the ability to direct their marketing towards them and in return, they would influence others,” said Benyoucef.