This is guest post by Dan Zarrella, a social media marketing consultant. You can follow him on Twitter here.
While some people have said that Digg has begun to lose its relevancy since the recent algorithmic changes, I believe it still represents an incredibly rich resource for studying social media and how stories and links spread throughout the web community. Once a link "goes popular" and is listed on Digg's homepage it is seen by many and perhaps even a majority of web geeks. Very often these readers have their own blogs, and if they like a story they may blog about it or link to it. This is why many webmasters yearn to be Dugg -- not for the first wave of traffic, which is often substantial but hard to retain, but for the viral wave of traffic and links that comes as a result.
Plenty of research has been done via Digg's API to study how stories make the homepage. We know the best time to submit to Digg, and we know the best categories and types of stories (e.g. top lists, how-tos and stories about Digg), but what has never been studied before is what happens after a link goes popular.
Digg's algorithm can be gamed just like any other, so we often see sub par stories hit the homepage, but low-quality stories rarely receive any substantial amount of links once they're made popular. It takes real, quality content to induce savvy social media users and bloggers to link to a story.
So I decided to study this effect: what types of stories get lots of links after going popular? I created a database of 33,000 of the 39,000 stories that made Digg's homepage in 2007 and, using Yahoo!'s API, I tracked how many external links each URL had pointing to it. Then I analyzed a number of factors that could influence the number of links a story gets and wrote a report with my findings.
I'm publishing my Link Attraction Factors (PDF) report exclusively for the first time here on ReadWriteWeb and I'd love to hear everyone's thoughts and comments. The report is available as a free PDF file.
Once I realized that the occurrence of certain words in a story's title or description can have a potentially large positive or negative effects on incoming link accumulation, I created two tools that leverage my database.
The Keyword Tool analyzes a specific word or phrase and returns the average number of links a story mentioning that keyword got in 2007. While the Title Check Tool analyzes an entire title string and shows you which words tend to increase links and which tend to decrease links. These tools can be used by webmasters or social media consultants to help them tweak their copy for optimal social media link attraction, or by other researchers looking to expand on my work.
I plan to expand some of this data for future reports, including by looking at social sites other than Digg and incoming link data sources other than Yahoo! Again, I'd love to hear all of your ideas in the comments.
Comments
Subscribe to comments for this post OR Subscribe to comments for all Read/WriteWeb posts
That is really interesting. I wrote my whole master thesis on the attention economy on social networks and explicitly in the blogosphere. The role of aggregators such as digg are of utmost important and your results are one more step on understanding attentiveness in the social web.
Posted by: robojiannis | February 20, 2008 11:30 AMYou can read the results of my research here and
download the whole thesis for free as pdf here
Good angle, Dan.
The links are much more valuable than the initial influx of traffic. The typical Digg visitor is passively looking for interesting stimulus, so their clicks are low value ones. But the linkbacks help you get PageRank, which helps you find the important prospects - the *searchers* who are actively looking for your product or service.
Chris
Posted by: Chris Goward | February 20, 2008 11:42 AMhttp://www.widerfunnel.com/blog
Dan,
What a neat tool is the 'Title Check Tool'!!
Well done. I am sure many will find it useful.
Hopefully you will keep the database updated on a regular basis :)
Posted by: efimor | February 20, 2008 12:04 PMThanks for the compliments!
Posted by: Dan Zarrella | February 20, 2008 12:09 PMefimor, i definitely plan to do some follow up research and update the DB.
Now I know why we didn't have a beer when I was in Boston a couple weeks back. You've been a busy dude. Well done - this is impressive, Dan.
Posted by: SEO Toronto | February 20, 2008 12:20 PMFYI - in the article i wrote yesterday and the one that Richard linked to - we received OVER 200 diggs in 24 hrs and did not hit the frontpage. In this same time period, ars was able to hit the frontpage with a story that at the time of their digging (22 hr into it) had 65 diggs less than mine.
hmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm
http://digg.com/tech_news/Can_You_Guess_Which_Categories_Make_The_Digg_Home_Page_Most
Posted by: allen stern | February 20, 2008 1:08 PMInteresting, thanks for the report. Those tools look really handy, too.
Thanks again,
Jack
Posted by: Jack | February 20, 2008 2:53 PMAllen,
I'm pretty sure upcoming digg stories compete with other stories _in their category_. Not against every other story on digg.
As of this comment
http://digg.com/tech_news/upcoming/most/
110 diggs: Getting Dugg and Going Viral: Link Attraction Factors
105 diggs: isoHunt Introduces New (Digg Like) Torrent Rating System
http://digg.com/apple/upcoming/most
99 diggs: Apple goes to Harpertown: a review of the new Mac Pro
The 99 digg story will promote before the 105 digg story since it's at the top of its category.
arn
Posted by: arn | February 20, 2008 4:10 PMThis is really interesting. Just today i posted my own analysis on digg related to the value of a single digg. By combining data from the digg API and the Alexa data service I managed to calculate what a single digg represents in terms of web traffic. Have a look:
http://anonymousprof.com/what%e2%80%99s-the-value-of-a-digg/
-AP
Posted by: Anonymous Prof | February 20, 2008 4:11 PMHow can you get -30% links (are you losing links that you once had)? That makes no sense to me?
Posted by: TVSpy | February 21, 2008 3:12 AMTVSpy, -30% links means that you are loosing out on 30% of the links you could get, basically you get 70 links instead of 100 from the average.
Good article and certainly a few surprisers in there (Travel!?) and 'breaking!' wasn't one of the top title words! I'm happy :) maybe people will stop using it
Posted by: Dudibob | February 21, 2008 6:15 AMwow, this comes at a good time for me. Great info, thanks Dan! I'll definitely use the Title Check Tool next time I Digg something.
Posted by: iwiLetter.com | February 21, 2008 6:31 AMVery Cool
Posted by: John Doe | February 21, 2008 10:10 PMDan's that's an awesome resource. I'm downloading the pdf and bookmarking the tools!
Thanks a lot :D
Posted by: Gab Goldenberg | February 22, 2008 9:00 AM