The 'Echoverse': New way to think about brand-consumer interactions

"You can't just be in your silo," says Roland T. Rust, Distinguished University Professor and David Bruce Smith Chair in Marketing at UMD's Robert H. Smith School of Business. "You have to manage all of your brand communications as a big system."

In addition to offering an unprecedented look at the interdependence of media, corporate communications, and information emanating from consumers, the study underscores how the importance of Twitter in brand management has exploded since 2010--even more than one might expect--and how the influence of other channels has waned. Consumer sentiment, for instance, a snapshot of how people feel about brands, is not nearly as important as it used to be. The study focused on the four top financial service firms from 2007 through 2013, a tumultuous period for that sector: Bank of America, Citibank, J.P. Morgan Chase and Wells Fargo.

The authors analyzed the volume and tone of messages in different media: articles in top newspapers; tweets over the period that emanated from or mentioned the banks; press releases; ad spending; and consumer sentiment. As a measure of business outcomes, the authors used deposits on a quarterly basis. They controlled for factors that would affect all banks simultaneously, such as general economic conditions. All told, the authors write, theirs is "one of the most comprehensive datasets in the brand communications literature."

To capture the "valence," or tone, of most messages, they used automated linguistic analysis. And to spotlight changes over time, they divided the dataset into two groups: 2007-2010 and 2011-2013. While people often talk loosely about an "echo chamber," this study offers an empirically rooted picture of how it actually works. Among the findings:

• Negativity feeds on itself. News articles with a negative tone lead to an increase in negatively toned tweets--and vice versa. Negative news articles breed more of the same, and negative tweets prompt more negative tweets. Such negative spirals lead to fewer deposits in banks. (Early in the period studied, positive feedback loops were more common. Later, the negative spirals became the norm. This may be partly due to the rise of online media, but traditional media also appears to feed off its own negativity more often these days.)
• Some parts of the echoverse influence each other equally. In other cases, influence goes one way. Traditional newspaper articles strongly influence word of mouth (Twitter, in this case); likewise, tweets affect the tone of news stories. In contrast, while Twitter influences consumer sentiment, consumer sentiment had little influence on what was said in Twitter. "This is in line with the opinion leadership role of the Twittersphere," the authors write. Companies issued more press releases when Twitter sentiment soured than when when consumer sentiment did.
• Press releases actually work. This surprised the authors, as press release are an often overlooked part of the echoverse. But positive press releases by companies could lift the tone of third-party tweets about the banks, and even lift customer deposits.
• Advertising bypasses the echoverse. Traditional ad buys did not affect traditional media coverage in a statistically significant way; nor did they affect tweet volume or tone, or consumer sentiment. Advertising did increase consumer deposits, however--consistent with existing literature suggesting advertising has a slow, long-term payoff.

The study also found that some banks had more effective social media strategies than others. Bank of America, the study found, was able to reduce the number of negative tweets mentioning the company, and even reduce negative news stories, by sending out more tweets. Wells Fargo could not. Unlike other banks, Bank of America, was also able to increase consumer deposits by ramping up its tweets. The authors suggested this is because Bank of America's main strategy was to send direct tweets to customers with concerns, whereas other banks were more likely to use Twitter as a kind of broadcast-advertising medium. (There was also some evidence that "high volume, consistent, moderately toned" tweets were more effective than more enthusiastic tweets that consumers may read as inauthentic.)

"Managers are used to a one-to-many model of communication," Rust says. "But more and more, they have to move to a one-to-one approach to be effective. On the other hand, consumer word of mouth used to spread one-to-one. More and more consumers are one-to-many in their brand communications."

The paper also implies that the considerable money companies spend on monitoring social media, including predictive analytics--predicted to reach $136 billion by 2020--is worth it.