DIGITAL COMMUNICATION REVOLUTION
Photo by Lane McLaughlin May 24, 2016
MIT's Sinan Aral discusses research on the power of social media
“Over the last 10 years, we’ve had a massive change in the way that we as human beings communicate with one another,” said Sinan Aral, a leading social media expert from the Massachusetts Institute of Technology (MIT).
“There’s been an explosion of digital social signals from our peers,” he continued. “Fifteen years ago, Facebook, Twitter, Yelp and every other social media technology that you can name didn’t even exist.”
Aral explored the implications of this evolution in human communication for business, health and public policy as the presenter of the University of Delaware’s sixth annual W.L. Gore Lecture Series in Management Science.
Aral serves as the David Austin Professor of Management at the MIT Sloan School of Management, where he holds a joint appointment in the IT and marketing groups and co-leads the Initiative on the Digital Economy.
His work utilizes huge swaths of data to understand how social media impacts organizational performance and consumer decision-making. He develops tools to measure and manage how information diffusion in social networks affects worker productivity, consumer demand and viral marketing.
“What is the implication for demand when individuals are buying products and then signaling to all of their friends that they bought that product or like that product?” Aral asked. “How does that demand cascade through the population?”
These social spillovers, he continued, are now accelerated by technology, amplifying people’s decisions and opinions by spreading them instantaneously to their networks.
“What does that do for the way that behaviors and opinions ebb and flow in society and ricochet from one person to the next?”
As part of one such study, Aral’s team worked with Facebook to study networks and determine which people in networks are most influential over others, and which are most susceptible to influence.
Aral explained that experimenting with this expansive data allows the team to make causal estimates on how an individual’s behavior influences the behavior of their friends.
“When you aggregate this at the level of the population, you can answer some pretty interesting questions,” he said.
In one example, the team was able to predict individuals’ susceptibility to others’ influence as a function of their relationship status. They found that single Facebook users were less susceptible to influence than those in a relationship or engaged, but that susceptibility dropped for married Facebook users. Most susceptible of all were those who reported their relationship status as “It’s Complicated.”
This example demonstrates the potential power of such information for marketers deciding where to focus their targeting efforts. If they target customers who are both highly influential and are friends with others who are highly susceptible to influence, they will “spread behavior change farther in the network,” Aral said.
“This will allow you to figure out who to message to maximize the diffusion of information, and to generate the most behavior change per dollar spent,” he added.
In another study, Aral’s team focused on how social networks influence not only behavior, but opinions as well. The team worked with a website to randomly manipulate items’ very first online ratings with either an upvote or a downvote.
They found that a single positive rating at the beginning of the rating cycle increased the mean of future ratings by 25 percent, and made a product 30 percent more likely to receive a score of 10.
“This is a huge number,” Aral said. “It essentially means that a small perturbation in the voting pattern at the very beginning totally changes the outcome” and “shoots you off into the stratosphere of high ratings.”
So what does this mean for business owners?
“If you’re a business, you want to encourage the customers who have had a positive experience to rate and rate early, not just because they’ll add one additional positive rating, but because it will influence all of the other ratings that come after it to be more positive,” Aral said.
Marketers should embrace this enhanced knowledge of social networks, Aral said, and this idea of a socially linked set of consumers. He described this mindset as the next logical step as marketing has developed from the single-message advertising of the 1980s through the segmentation of the ’90s and personalization of the ’00s.
“If your institution or organization is focused on segmentation, you’re about two decades or more behind,” he said.
And, he added, this information is important not just for commerce, but for social movements, health behavior and public policy as well.
“Does social media and the signals that we get from social media help us to get out in the square and protest, or get out to the polls and vote?” Aral asked. “Does Fitbit affect the way that your health behavior is spilling over into the health behavior of your friends?”
Similarly, public health campaigns that encourage individuals to quit smoking or exercise more could use this data to maximize their reach.
“How should I design a public policy intervention that is targeted at groups of connected individuals rather than a bunch of people that we assume are independent and unconnected from each other?” Aral asked.
“Maybe I can save a lot of money and get the same outcome in terms of behavior change in society if I take into account all of these social spillovers.”
This will remain a hot topic of statistical inference and experimental design, Aral said, as it can help “generate more large-scale behavior change from the same amount of resources.”
About the Gore Lecture
“It’s a huge honor to give this lecture with so many distinguished people having given it before,” Aral said of the W.L. Gore Lecture Series in Management Science.
The Gore Lecture, presented by UD’s Alfred Lerner College of Business and Economics, is sponsored by an endowment from the Gore family.
This lecture series features experts in the application of probability, statistics and experimental design to decision-making, including applications in academia, business, government, engineering and medicine.
The series recognizes the key role that the fields of probability, statistics, and experimental design have played in the success of W.L. Gore and Associates Inc.