February 28, 2012 | Research FeatureEvent Highlights

Why Do You Tweet?

New research from Olivier Toubia shows why people use Twitter and points to how firms should capitalize on the social media platform.

Topics: Marketing

Professor Olivier Toubia knows social media: his research often investigates the way online networking tools can be leveraged for marketing. But when it came to Twitter, the social media platform that limits users to 140-character posts, he says, “I was puzzled by its popularity. Most users have no financial incentive to post or follow others, so what is the basic motivation for being active? Before companies can harness Twitter, they have to understand why people use it.”

Toubia was also intrigued by the asymmetry of Twitter as compared with so many other social media platforms, because reciprocity isn’t a given: it is possible to follow a user without that user following you in return. This feature makes Twitter an ideal environment for studying social status. “Because it is possible to have large differences between the number of followers you have and the number of people you follow,” Toubia explains, “you can see status emerge.” Along the same lines, Twitter also allows researchers to observe status in a way they could not in the past. “It’s very hard for a researcher to make some kids more popular at school and see how they react,” Toubia notes. “These types of manipulation that are impossible to do in a face-to-face setting are now possible in online settings, giving us new tools to study some old and fundamental questions.”

Toubia points to the many press reports of Twitter’s usefulness as a tool of self-promotion, in which users’ numbers of followers has become a form of social currency, suggesting status as a motive for being active on Twitter.

“If Twitter is useful mostly for broadcasting and sharing information, as it was initially positioned, then, like a newspaper, the bigger the audience the more value you get out of printing, and you would expect people to post more as they get more followers,” Toubia says. “But if Twitter is about status and bragging rights, and being able to say you have X number of followers, then any new follower is going to have diminishing returns,” he continues. “Your 1001st follower is not going to give you as much pleasure as your 101st follower and you won’t work as hard (i.e., post as much) to attract this new person.”

Toubia worked with Andrew Stephen, PhD ’09, now of the University of Pittsburgh, to look for evidence of what motivates Twitter users. Is it intrinsic motivation — an internal desire not influenced by external rewards like money — to broadcast or share information with friends and the world at large? Or is it simply that users want to build social status?

To answer these questions, the researchers needed to track users’ responses to increases in their number of followers. But understanding the impact of followers on user behavior is something of a chicken-or-egg problem. The researchers couldn’t simply observe the correlation between number of followers and posting frequency because doing so doesn’t reveal whether an increase in followers prompts users to post more often, or whether posting more often prompts more people to follow those doing the posting.

To solve this problem, the researchers made one of the variables (the number of followers) bigger, and tracked the changes in the other variable (users’ frequency of posting) and compared this behavior to a control group. Taking care to conform to Twitter’s Terms of Service, and with the help of undergraduate research assistants, Toubia and Stephen created a number of artificial users. The researchers assigned these “fakes” to follow real users, increasing the number of fakes following these users over time, then noted whether those users, in response to their growing popularity, posted more frequently.

While on average all the tracked users posted more often compared to a control group of users, the details are telling. For users who had very few followers to begin with, becoming more popular didn’t change anything. “These are probably people who come to Twitter, create an account, post twice, and never come back,” Toubia explains. Increasing followers for super-users didn’t have much of an impact, either. “Getting 100 new followers is not a big deal for someone with 10,000 followers,” Toubia says.

But users who had a moderate number of followers — from between 13 and 26 to start with — posted more frequently as they gained followers. “That behavior is consistent with the idea that the more followers I have, the more value I derive from posting, so I post more,” Toubia says.

Users who already had a somewhat higher number of followers — between 62 and 245 — actually posted less as they became more popular. “This behavior is consistent with the idea that these users are on Twitter to earn status, and that once they reach a certain status — reflected by their number of followers — their motivation to post is reduced.” This suggests that some users have an intrinsic motivation primarily to broadcast to the world, while another group of users has a more image-conscious, status-seeking status motivation.

The researchers also created a model that quantifies how much value Twitter users receive from different sources of motivation and allows for predicting behavior should Twitter become stable (i.e., few or no new users join the social network). “What will happen as Twitter matures? Will people post more or less? Our model suggests that as the network matures and becomes more stable, it will hold less attraction for those who view it as a marker of status because there will be fewer new followers to gain,” Toubia says.

As a result, non-commercial users may become increasingly passive consumers of content. “That is consistent with what we have seen over the last two years, with Twitter becoming more of a broadcast channel. You have the super users, like Kim Kardashian, who have millions of followers and become like media channels,” Toubia says. “Then there are all the other users who are more passive consumers of content from superusers. There are many channels, and you follow the ones you want to subscribe to. Users become consumers rather than content creators.”

Twitter has recognized the shift. “Previously, Twitter’s homepage positioned it as a way to share information,” Toubia notes. “Today, it is positioned as a way to connect with the latest news that you find interesting. The company emphasizes that ‘you don’t have to tweet to enjoy Twitter.’”

For now, firms that want to make the most of Twitter should recognize broadcasting and social status as two key motives for posting. “If firms want to engage with people on Twitter, they should engage in a way that recognizes those people’s motivations.”

Olivier Toubia is professor of marketing at Columbia Business School.

Read the Research

Olivier Toubia, Andrew Stephen

"Intrinsic vs. Image-Related Utility in Social Media: Why Do People Contribute Content to Twitter?"

View abstract/citation  Download PDF  

Have You Read Columbia Business?

Columbia Business School’s alumni magazine connects alumni with each other and the School; celebrates alumni milestones and accomplishments; and chronicles the impact of Columbia Business School alumni, faculty members, and students on the global business landscape.

Read Columbia Business>