Over a million developers have joined DZone.

Research revisits what makes us retweet

DZone 's Guide to

Research revisits what makes us retweet

· ·
Free Resource

Twitter is undoubtedly a popular channel for sharing information.  As such there have been an increasing number of studies looking at just what kind of information tends to get shared most frequently.  For instance, back in 2012 some researchers from MIT explored the topic and came up with 10 tips to help you secure more retweets, including sharing success, helping people, and even asking for the retweet.

Fellow MIT researchers then followed this up with an attempt to predict whether a tweet would be shared or not.  To showcase their model, they created a website where you could see the latest tweets from various luminaries together with a prediction of how many retweets it would receive.

Researchers from W.P Carey School of Business then analyzed the kind of network relationships were more prone to retweet our stuff.  Intuitively one would imagine that the stronger your relationship, the greater the chance of them retweeting your content.  Alas, the study found that wasn’t the case, and it was more likely that weaker ties would share your stuff.

This finding has been explored by a new study published jointly by researchers from Utah State University and IBM’s Almaden research centre, and they believe they’ve struck on a method that can increase the chances of a retweet by up to 680%.

Central to their theory was that people are more likely to tweet on certain topics at certain times of the day, and that targeting them accordingly is the best way to secure the retweet.  All of which is pretty straightforward.

The team used an algorithm to draw up some data around their target users, including their level of retweeting, the ratio of followers/friends and the time they usually tweeted.  They found that the algorithm was pretty good at picking those who might share their update, with a 13% success rate when those users were included in the tweet, against a 2% ratio for the control group.

When this was combined to include messaging those people at times of high activity, this ratio improved to 19.3%.  Suffice to say, the app isn’t publicly available, but it doesn’t seem too much of a stretch for the various Twitter developers to build a tool that can perform this function for users, and for such an app to be extremely popular amongst marketing types.

Original post

Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}