Researcher cracks the code on what makes a tweet popular

Twitter Happiness

Twitter has become an essential part of the small business marketing playbook, but it remains a difficult platform to truly get your arms around. Some tweets will go gangbusters on the service, replied-to and retweeted for hours. Others will land with a thud, quickly drowned out by the chatter of the masses and unceremoniously forgotten.

Wouldn't it be great if you could know whether a tweet you sent was going to be a hit or a miss?

Now you can, according to Tauhid Zaman, an Assistant Professor at the MIT Sloan School of Management who researches social media analytics.

Specifically, Zaman has researched the curious phenomenon of the retweet. To a large extent, retweets are the lifeblood of Twitter, as other users re-broadcast something you've tweeted to their followers. Done right, a solid tweet can create a chain reaction of retweets as friends of friends of friends pick up your original message and give it an exponentially wider audience than you ever could alone.

Zaman and his co-researchers looked into the variables that made tweets popular by analyzing a collection of 52 tweets from a variety of major Twitter accounts belonging to everyone ranging from Garry Shandling to Ann Coulter. The team then built a model to predict how many retweets any given tweet would receive. All the data has been collected on Twouija for you to check out firsthand. Select a Twitter handle, then examine the blue line, which shows how many retweets a message received over time. Click at any point on the blue line and you'll see a red bar appear above it: That's the total number of retweets the post is expected to ultimately receive.

A sample graph of a post's retweets via Twouija.

How is that number calculated? Simple, says Zaman, because the total number of retweets a post ultimately gets is largely determined by the number it receives in the first 10 minutes of its life. "In the first 10 minutes you'll get about 10 percent of your retweets," says Zaman. "It's the same for everybody, whether you're Barack Obama or a small-time blogger."

In other words, it doesn't matter if you have a million followers or just 1000: Those first few minutes are crucial for your tweet's survival, and even the Prez can drop a DOA tweet once in awhile despite his massive following if too few people take notice right away.

Zaman wants to use the research to develop a more sophisticated way of predicting and ranking the popularity of tweets and Twitter users that would go beyond what simply counting followers can do. "It's a whole new way to rank content in social media." Zaman hasn't set a timeline, but eventually the Twouija site will be opened up so that anyone can analyze any tweet to get a prediction for how popular it will become in real time.

For business users, the ramifications of Zaman's research are intriguing. Traditionally, Twitter marketers will compose a tweet, send it out, and wait for hours to see if it managed to gather a critical mass. Zaman says that's folly: If you aren't seeing a near-immediate pickup of your tweets, your best move is to try again with a new one rather than hope that the original will eventually find its mark. "It doesn't matter if you have a lot of followers or a few. If you don't hit a critical mass in the first few minutes, the chances are very slim that your tweet will have broad appeal," says Zaman.

Zaman also sees applications for those who use Twitter to advertise. Tweets that show early promise using Zaman's algorithm might be marked to be promoted, comfortable in the knowledge that they'll gain an even greater audience through other users jumping on the bandwagon.

To comment on this article and other PCWorld content, visit our Facebook page or our Twitter feed.
Shop Tech Products at Amazon
Notice to our Readers
We're now using social media to take your comments and feedback. Learn more about this here.