via @predictive_users
April 21, 2011 1:56 PM   Subscribe

Is there any way to identify social media users who are consistently sharing viral content before it goes viral?

For my own amusement, I am interested if is possible to identify users that seem to pick up on material that has viral potential early on. I'm not really interested in power users who, once they share it, will cause it to go viral. I'm looking for average joes who seem to have zeroed in on what people think is funny/relevant. Especially if the material is found firsthand, or from a known contact vs on a site like Metafilter or Digg. I wonder if these people are right in the middle of our collective sense of humor or curiosity. It doesn't matter if they are popular or well-followed, just that they are early and accurate.

I know this question is kind of obtuse and I apologize for using the word viral enough to be annoying. Its just been on my mind and I would like to get other people's thoughts on it.
posted by halseyaa to Computers & Internet (8 answers total) 7 users marked this as a favorite
I can't say that there's necessarily a way to do it, but this is how I'd go about it:

1) Cache twitter (Some content analysis researchers have done this, though I'm not sure if it's available to the public. Looks like there's some search engines here.)
2) Pick a couple of cases across disparate media- articles, youtube videos, etc.
3) Track chronologically the links and retweets of your chosen cases
4) From among your earlier occurrences of links, exclude users who were frequently retweeted
5) Repeat a bajillion times until you find overlap

Sorry to turn this into a research project. On the plus side, the cynical academic in me thinks you could probably get a grant for this, considering how hot social media research is right now.
posted by libertypie at 2:08 PM on April 21, 2011

I mean, sure. But only by, as libertypie suggests, putting a heck of a lot of work into it ahead of time. There isn't going to be any part of any site's UI that's going to give you this sort of data. You'd have to find some way of exporting data of who has liked/tweeted/whatevered stuff and then run comparisons.
posted by valkyryn at 2:27 PM on April 21, 2011

My work intersects with this problem somewhat, though not with Twitter. What you're talking about is the Holy Grail for marketers looking at social networks. If you can figure that out, you'll be very, very rich.

Take a look at stuff like Big Sheets to understand how large this problem is in scope.
posted by mkultra at 2:31 PM on April 21, 2011 [2 favorites]

Not to toot my own horn, but I seem to link to quite a few things on Facebook that I find floating out on the web, then the next day or several days later I see the same things posted on major websites. I think it comes from reading so many disparate sites each day - everything from pop culture to science to video games to religion.

I just pick things that amuse me (videos, user-created images, articles, etc.) and Share them so other people can enjoy them too.

Look through your Facebook friends and see if you have any like me who share for the sheer joy of sharing.
posted by tacodave at 3:18 PM on April 21, 2011

When the panic about being closed was in full force, I read an entry on a blog lamenting the possible loss and explaining one way they used it (sorry, no idea which blog). They took all of the things they'd linked to over the past year, then looked to see which user was the first one to save that link (the original, not the blog post) to delicious. Many users showed up several times, so they started following them in delicious and found a treasure trove of early links to things about to hit it big on the web.

Really, you could do that pretty easily with all of the stuff that shows up on the hotlist for a few weeks.
posted by soelo at 3:29 PM on April 21, 2011 [1 favorite]

Its just been on my mind and I would like to get other people's thoughts on it.

As mkultra says, this is not an easy thing to do. Anyone who actually can consistently identify trends before they become well established has a long and very lucrative career waiting for them. It's kind of like asking if there is a good way to identify stocks which are going to go way up before they go way up. Maybe there is (see Buffett, Warren) or maybe if you take a large enough population some people are bound to have a long streak of luck. But it's probably not something you're going to stumble across on a rainy afternoon.

The novel Bellwether by Connie Willis is actually about exactly the type of thing you're asking about. It's fiction, alas, but you might be interested in it.
posted by Justinian at 4:03 PM on April 21, 2011 [1 favorite]

Backtype shows you influencers of interesting people.
posted by harmfulray at 4:55 PM on April 21, 2011

I just saw this. It doesn't directly answer your question, but you might find it interesting anyway. Apparently, a big factor whether or not "Tweeter 0" is followed by a celebrity. I guess that's also kind of what Backtype is getting at as well.
posted by natabat at 2:47 PM on April 22, 2011

« Older Help, I'm a Platonist!   |   Why is entropy necessary for there to be memory? Newer »
This thread is closed to new comments.