How Techmeme Might Work
It is Saturday and it is time for tech news to turn its gaze inwards once again. Today, Jason Calacanis has once again put out a juicy piece of link bait. This time, it’s about Techmeme. Ignoring the rest of it (he does make a certain amount of sense), I wanted to instead offer up some conjecture as to how Techmeme actually functions.
Below is little more than idle speculation, but it is pretty consistent with my experience with Techmeme’s news tracking.
Techmeme
First, you need to start with an initial list of arbitrarily chosen sources. I say arbitrary because the bot, scrapper, algorithm or what have you needed some place to start its journey. This type of seeding is likely how sites like the New York Times, Scripting, Techcrunch, Digg and the generic tech media got into the system.
Alternatively, if you don’t like the idea of an initial seed set of sites — you can assume that the algorithm pulled in a list of heavily linked sites that deal with technology news. I tend to believe the initial seed theory though, but since it has been years since it would matter, lets move on.
After the bot, scrapper, algorithm knows where to look, it trolls the Internet for cross-linking behavior.
An example would be that Techcrunch writes an article about a, “New Google Widget BETA With Collaborative Tagging.” Within some set time span, Digg, Dave and the New York Times links to the article. Since so many sites with a high level of Techmeme authority linked to an article, it has to be important. Thus, it gets a headline slot.
The question then becomes, “How do new sites make the cut?”
Well, you have to make it into the discussion index first. Once again, I believe links determine this. If you write an article that gets linked to by sites already in the index, you are deemed to be important enough to be added to the discussion. That would explain why random stories that have made Digg sometimes make it into the mix.
How do sites become headlines without links? Things become a little more confusing at this point. I think after you’ve made enough headlines, the bot just scrapes your page on occasion for content to start discussions. Whether Gabe acts as an editor in these cases or the algorithm is just very sophisticated, I can’t really tell. I tend to subscribe to the “light editorial” model.
Web 2.0 Roundup
That seems to be the long and short of how Techmeme works. I think there is some arbitration and editorial influence to make sure some of the more obvious forms of gaming are kept out of the mix, but at the end of the day — Techmeme is a link game. If anyone has other theories, feel free to drop me a line or leave a message here.
[Be sure to subscribe to the RSS feed before leaving. Photo Credit]
