Dan Zarrella: How ReTweets Spread: The Epidemiology of Viral Messaging on Twitter
Ezra Butler
· 11 months ago
Very interesting. How does the algorithm deal with the erasing of previous people's names due to the space constraint? I especially like your proof of the social proof.
Jessica Routier, IACEZ
· 11 months ago
Interesting! Just more proof that tweeting is worth the time! (Who cares that it's fun?!?! :) )
Troy Peterson
· 11 months ago
LOVE this line of thinking. If you need help with content analysis. I'd be willing to participate. There is going to be a very real value in having data to determine what TYPES of tweets are most successfully retweeted. I'd also like to see how requests for R/T's fail.
Very interesting, very cool!
Troy
Chris Lockwood
· 11 months ago
I don't think most people realize that a retweet is not seen by many new people, mostly just those who are already following the person you are quoting.
In other words if @bob says this: "RT @joe something cool" most of bob's followers will NOT see that, just those who are already following both bob and joe. This is the Twitter default that each user has to change in their account if they want to see more of these.
Dan Zarrella
· 11 months ago
@chris that seems to be a common misconception today. Only tweets that begin with @ are hidden, not tweets that begin with rt or retweet and then the @.
Peopleshark
· 11 months ago
Very curious about job tweets...wondering if job opportunities are shared more widely among networks than other kinds of tweets. Your analysis is very helpful. Thank you!
Nathan Ketsdever
· 11 months ago
Chris' comment may rest on flawed assumptions. The repeat in the same friend group (social graph, tribe, whatever) may actually boost the value, because those receiving the tweet know who Dan Z is or whomever the original tweeter is.
Jeremy Mandle
· 11 months ago
Great work Dan. I'm wondering how much your queries take into account the different RT, ReTweet formats used?
For example: tweet (via @twittername) I tend to use this most often as it is a bit more "readable", and is the default retweet format in Tweetie.
Applications like Twhirl all you to completely customize your retweet format.
It has been awhile since I've studied the XML anatomy of a tweet BTW.
Sean Carmody
· 11 months ago
@Dan: I wasn't aware that any @ messages were hidden (I see plenty in my stream). When/where are they hidden.
P.S. Love the RT analysis.
Jim Tobin at Ignite Social Med
· 11 months ago
Good stuff Dan. Love seeing some sociology and some hard data applied to this stuff. Keep it coming.
Ironically, I found this by Tweet and I'm going to retweet it.
~Jim
Erica DeWolf
· 11 months ago
Thanks for this research. It will be very effective in proving to people the power of twitter
Brian Carter
· 11 months ago
Two thoughts about your retweet mapper findings:
1. What makes retweets more likely to be retweeted? (beyond that first retweet)
2. Does it make sense to separate retweets that contain links vs those that don't? I suspect sharing a link makes retweeting more likely BUT retweets of just ideas without links are also valuable in a different way- what makes them retweet-worthy?
Another factor is how some Retweets are solicited. Many times people use DM's to request Retweets, either directly or indirectly. This would be tough to factor into the retweet equation. I've received some very aggressive requests and they often end up being retweeted by others quite a bit, even if half-heartedly.
@toddlucier
· 9 months ago
Interesting research Dan. I see a problem emerging with Tweetbacks that I think you need to be aware of in your research.
By definition, Tweetbacks that are merely RT's of previously posted content or links - as featured in the comment section of a blog are clearly Spam.
Tweetbacks inherently give those who retweet posts significant profile on blogs that feature tweetbacks.
This benefit of visibility for simply pressing a button and associating oneself with a successful blog is the sort viral positive feedback loop we can do without.
For the blogger, the RT's get their blog more exposure, for the Retweeter the RT gets them exposure for their twitter brand. . . in the process blogs like yours get filled with useless links that do not contribute to the discussion and in fact are nothing more than spam.
It's a lazy person's way of a Twit pulling on the bloggers coattails.
What makes RT comments on a blog any different from Spam comments like - "Hey, nice post."
I look forward to your response. The use of tweetbacks clearly has consequences for your statistical analysis and use of such tools clearly needs sober second thought.
After all, what's to slow someone from following peeps who's blogs feature tweetbacks and constantly RTing their content to get followers. Seems to me the reward of featuring a tweet on a blog post for a tweet isn't desirable.
taa
· 8 months ago
wow - you're one really smart guy! You truly are "a social & viral marketing scientist" and I look forward to following you on twitter shortly!
All Comments
I especially like your proof of the social proof.
Very interesting, very cool!
Troy
In other words if @bob says this: "RT @joe something cool" most of bob's followers will NOT see that, just those who are already following both bob and joe. This is the Twitter default that each user has to change in their account if they want to see more of these.
For example:
tweet (via @twittername)
I tend to use this most often as it is a bit more "readable", and is the default retweet format in Tweetie.
Applications like Twhirl all you to completely customize your retweet format.
It has been awhile since I've studied the XML anatomy of a tweet BTW.
P.S. Love the RT analysis.
Ironically, I found this by Tweet and I'm going to retweet it.
~Jim
1. What makes retweets more likely to be retweeted? (beyond that first retweet)
2. Does it make sense to separate retweets that contain links vs those that don't? I suspect sharing a link makes retweeting more likely BUT retweets of just ideas without links are also valuable in a different way- what makes them retweet-worthy?
People must understand how oiwerful Twitter is.
Thanks for this info!
Regards,
André T
I see a problem emerging with Tweetbacks that I think you need to be aware of in your research.
By definition, Tweetbacks that are merely RT's of previously posted content or links - as featured in the comment section of a blog are clearly Spam.
http://www.tourismkeys.ca/blog/2009/02/is-twitt...
Tweetbacks inherently give those who retweet posts significant profile on blogs that feature tweetbacks.
This benefit of visibility for simply pressing a button and associating oneself with a successful blog is the sort viral positive feedback loop we can do without.
For the blogger, the RT's get their blog more exposure, for the Retweeter the RT gets them exposure for their twitter brand. . . in the process blogs like yours get filled with useless links that do not contribute to the discussion and in fact are nothing more than spam.
It's a lazy person's way of a Twit pulling on the bloggers coattails.
What makes RT comments on a blog any different from Spam comments like - "Hey, nice post."
I look forward to your response. The use of tweetbacks clearly has consequences for your statistical analysis and use of such tools clearly needs sober second thought.
After all, what's to slow someone from following peeps who's blogs feature tweetbacks and constantly RTing their content to get followers. Seems to me the reward of featuring a tweet on a blog post for a tweet isn't desirable.
http://twitter.com/720media
1. Filtering- only the most retweetable messages propagate deeply.
2. Network Topology- at higher RT depth, the message accesses a more globally connected social network.