The Community Engine Blog
News, tools, and analysis for innovating in the information economy
« I quit using my feedreader for a month, but I've started again | | How much Internet can small businesses digest? »
Mass Conversation Tracking Just Doesn't Work
Sites that tell you they will help you follow conversations just don't have the coverage of the web to do so. A practical solution is to focus on a small community and use the larger services to catch general mentions.
Sections: Education Emerging Practice
We're in the third week of the latest blogging bootcamp, and we're just about to cover how to track conversations. In the process, I've come to the conclusion that mass conversation tracking just doesn't work. It's pretty easy to see why.
Just consider technorati. They track 26.3 million sites and 1.9 billion links. That's less than 75 links per site. In a good month of blogging, I can generate that on one site. So, technorati's coverage is not adequate to the task of tracking every link. The same is true of icerocket and Google blog search.
So, what's a person to do? I think there are two practical solutions:
- One is to use TalkDigger as first brought to my attention by John Tropea in the Corante newtork. It pulls together results from a number of engines and gives them to you in a nice RSS feed. That doesn't solve the coverage issue, but it does mitigate it to the extent possible. If you keep the search granular enough, you'll at least catch new people noticing your site. BTW, this is functionality Andy Seidl was wowing me with at MyST Technology a year and a half ago, and now you can get it for free.
- The other is to focus on a group with whom you interact regularly and then just follow them closely. This is clearly how it worked in the early days of blogging with people like Dave Winer and Doc Searls. It's also a little how it works with places like the Corante Network or the BIT320 Remix I ran in a class last Fall. Participants in the current bootcamp will likely set up their own suprglu sites to achieve the same effect in their teams of five.
Bud posted this on January 24, 2006
Trackback Pings
TrackBack URL for this entry:
http://thecommunityengine.com/cgi-sys/cgiwrap/fpgibson/thecommunityengine.com/mt/mt-tb.cgi/6606
Comments
Hi Mr. Gibson,
You are completely right. They are not that efficient. But even if they were, you would have another problem: how to present all that information in an efficient way? This is a problem, a big one.
Another problem is that rare are the people that link that often. I think that we will have to create a sort of “linking-education”, especially to new web content publisher (and there is a lot new with the recent coming of Blogs). They have to know why they should link, why it’s good and what it changes.
I am happy to see that you seem to like Talk Digger, and thanks for its mention. It is sure that it is really not perfect, but I hope coming with something more useful in the future: something that help people to “see” these “conversations”.
Thanks,
Salutations,
Fred
Posted by: Fred ![[TypeKey Profile Page]](http://thecommunityengine.com/home/nav-commenters.gif)
at January 25, 2006 01:42 PM
Great article, Bud, and very true. Monitoring public conversations that occur via blogs and other types of conversational media is a challenge.
The best approach I've found so far is to master the art of specifying a strong collection of search feeds -- often the same queries across several feed-aggregation resources to increase coverage. I create a folder of these feeds for every client or project.
It's not perfect, but it's pretty damn good -- and fast. Plus, it pulls in content not just from blogs but from many sites and mainstream media outlets that offer feeds.
I complement this with TalkDigger and other tools/strategies on occasion.
But I'm always open to hearing other solutions.
- Amy Gahran
RightConversation.com
Contentious.com
Posted by: Amy Gahran at January 27, 2006 05:47 PM
