HP Invents a Formula to Predict Twitter Popularity
Over the last few years, Twitter has played a pivotal role in content marketing. Its wide reach and large user base gives marketers the potential to reach thousands, if not millions of people, while the microblogging site's intuitive user interface and features allows content to be shared quickly and easily. Additionally, since Twitter shows the number of retweets and replies that a post has received, it's easy for users to track how each tweet performs.
It's no wonder then that most marketers nowadays are competing to come up with the most creative tweets and posts that can grab the attention of users and garner the most replies and retweets as possible.
If only there was a special formula that can determine whether or not a post will perform well in the Twitterverse—some sort of pattern that marketers and content creators can use to ensure that their articles will garner maximum clicks and retweets. Can such a thing even exist?
Well, according to the guys at HP Labs, this formula DOES exist, and they were the ones that came up with it. Earlier this week, HP revealed the findings of a study that aims to forecast popularity on Twitter for a news article before it is published. They claim that they were able to predict how well a post will fare with up to 84% accuracy.
How They Did It
The study, entitled "The Pulse of News in Social Media: Forecasting Popularity," was done by Sitaram Asur, Roja Bandari, and Bernardo Huberman. They gathered over 40,000 news articles published over nine days in August 2011, and analyzed their Twitter popularity using four factors:
1. The news source that posts the article – "Does it make a difference who publishes a news article?" That is, will a post from say, the New York Times be retweeted more than a post from an unknown news source?
2. The category of the news article – "Does the category of news affect its popularity within a social network?" For instance, will a tech post do better than an article about health?
3. Whether the language of the post is objective or emotional – "Do readers prefer factual statements or do they favor personal tone and emotionally charged language?" Are users more likely to share a post that's about facts over an article that relates to their emotions?
4. Whether famous brands or celebrities are mentioned in the post – "Does it make a difference whether or not famous names are mentioned in the article?" Would a post mentioning Steve Jobs be clicked and shared more just because it includes his name?
The researchers scored each article using the abovementioned factors. In order to predict its popularity, "the number of tweets the article would receive was calculated using standard statistical models."
According to HP, the method above was able to predict (with 84% accuracy) how often an article will be tweeted. Apparently, "stories that mention celebrities, come from credible sources, and belong to popular categories of news (e.g. technology) are more likely to generate tweets."
However, the researchers discovered that the tone or language (whether an article is objective or emotional) "had a very minor influence on its distribution on Twitter, suggesting that "link bait" headlines and hysterical adjectives don't propagate messages any further than straightforward reporting."
While predicting a tweet’s popularity can't be an exact science, it wouldn't hurt to test HP's findings yourself. Try examining your tweets and see if posts about tech or articles mentioning famous people are indeed more popular. If you're more of a content curator, see if posts coming from credible news sources get retweeted the most.
However, it's also important to remember that there is still a lot more to learn about social media, so you shouldn't let one study rule over how you create and publish content. Simply use HP's method as a guiding factor to how you write your posts, but you shouldn't follow it to the T.
Ultimately, it's people and not formulas that make great content, and while HP's study is interesting and a good thing to keep in mind, don't forget that overall value and substance are the main reasons that an article gets retweeted. And no amount of name dropping and technology can make a post go viral, if it isn't well-written to begin with.
Image credit: Rosaura Ochoa on Flickr.