Big Data at the Box Office
“Back at work and recovering from #avatar—fantastic movie!”
“Wow! I wanna see ‘the lovely bones’!”
Two types of tweets about two different movies. But from a business standpoint, which type of tweet carries more weight in affecting a product’s sales revenues?
Probably the second one, according to Huaxia Rui, assistant professor of computers and information systems at the Simon School of Business.
Rui and two fellow researchers analyzed the impact of four million tweets on box office sales for 63 movies. So-called “intention tweets” from people who hadn’t even seen the movies appeared to have a greater effect than “positive tweets” from people who had actually seen them.
The results of their study suggest that online chatter really does matter in affecting sales. Business managers could glean important clues about the popularity of their products and even forecast future sales through careful analysis of Twitter traffic.
Word of mouth has always been regarded as a major influence on whether a product will be a big seller or not, Rui says. With the advent of Twitter and other social media networks, huge numbers of word of mouth messages are easily accessible to researchers and business analysts alike, providing “real time” indications of consumer preferences and reactions. Twitter is especially fruitful because it allows researchers to extract the number of followers each author has.
From June 2009 to February 2010, Rui and his colleagues used a computer program to query Twitter every hour for messages mentioning 63 different movies. They filtered out institutional messages; they used machine learning algorithms to classify tweets into one of four categories: those that showed an author’s intention to see a movie and ones that gave a positive, negative, or neutral opinion about a movie already seen.
They then used a dynamic panel data model to measure the effect of this word of mouth traffic on weekly box office sales.
They found that the more chatter there is about a movie, the higher its sales revenues, especially when there is a relatively high ratio of tweets from authors with a high number of followers (for example, 400 or more).
Not surprisingly, positive tweets boosted revenues, and negative tweets decreased them—casting doubt on the old saying that “any publicity is good publicity”!
The most surprising finding, at least on the surface, is that “intention” tweets from people who had not yet seen a movie appeared to have an even stronger impact on revenues than positive tweets from people who had actually seen one.
Rui suggests this is because of the dual effect of intention tweets: These tweets are a clear indication that their authors intend to see a particular movie, and the tweets not only make their followers aware of the movie but possibly influence them to see it as well. That makes “intention tweets” far more valuable in attempting to forecast future sales.
How might a savvy businessman use this kind of information?
Imagine you’re the manager of a retail store, and it is two weeks before Black Friday. If you are scanning Twitter and detect a surge in “intention” tweets showing an interest in one of your products, “That could be useful for determining your staffing and inventory,” Rui notes.
With more people using smart phones, tweets even reveal geographic location, which could narrow such staffing and inventory decisions to single regions—even individual stores.
“It sounds futuristic because nobody has done this. But I think it could be useful in the future,” Rui says.
Consumers could benefit as well.
Businesses that monitor social media, for example, will likely be more responsive in addressing complaints reflected in negative tweeting, precisely because the tweets will be visible to so many other potential customers.
Rui is actually working on a system called twitter sensor to give consumers even more power to check how well companies are treating their customers based on people’s discussions on Twitter. And customers may be less likely to find long lines or empty shelves on Black Friday if their local stores have done their Twitter “homework” in advance.
Rui and two fellow researchers analyzed the impact of four million tweets on box office sales for 63 movies. So-called intention tweets from people who hadn't even seen the movies appeared to have a greater effect than positive tweets from people who had actually seen them.