top of page


MIT Sloan Management Review
December 3, 2018

Nordstrom, the Seattle-based retailer, had a memorable 2017. In early February, Donald Trump, then the newly elected U.S. president, took to Twitter to berate Nordstrom for dropping the Ivanka Trump clothing line, complaining that the company had treated his daughter “so unfairly … terrible!” The tweet set off a powerful reaction in social media. Our research showed the number of weekly mentions of the Nordstrom brand on Twitter and other sites surged by 1,700%, while the tone of those conversations (as measured using natural-language processing, which interprets meaning from adjacent words and context) swung sharply from positive to negative.1 However, in off-line conversations (measured via surveys), the sentiment stayed positive. Amidst these mixed signals, Nordstrom rolled through the 2017 holiday season with a 2.5% sales increase over the prior year.

Divergent conversations about brands are fairly common — and not only for brands that get caught up in controversies.2 Indeed, we studied more than 500 leading consumer brands and found that in most cases there was little correlation between what consumers said about the brands online and what they said off-line, even though both streams of conversation can have big effects on a company’s sales.3

Marketers have long recognized word of mouth as a powerful force affecting how well products perform. Since the advent of Twitter and Facebook, some people now think of social media as “word of mouth on steroids” — the conversation that represents what consumers are saying.4 Yet we found that online and off-line conversations matter for different reasons.

Most studies on social media marketing effectiveness have looked at how brand engagement on specific platforms such as Facebook or Twitter (for example, the likes, shares, retweets, and comments) responds to marketing initiatives as opposed to considering the social ecosystem as a whole. There is little research looking at off-line conversations — those that occur face-to-face at the office watercooler, over the kitchen table, or at a health club — because of the difficulty and cost of measuring them. However, we addressed that challenge by asking selected consumers to recall the product and service categories and brands they talked about the day before, including whether the brand conversations were positive or negative.

Deriving Value From Conversations About Your Brand: Resources and Tips
bottom of page