Techmeme Reverse Engineering Report
11:05 am June 2nd, 2007 by Sal Cangeloso
Scoble had an interesting post a few days ago called “Techmeme: The Anti-Linking Engine“. It seems to have stayed under the radar of most people but it caught my attention, not because I have noticed the sort of trends that Scoble has, but because I am really interested in Techmeme (I’m not alone here) and because this is the sort of over-thinking that I tend to do about a lot of things.
Gabe Rivera, the brain behind Techmeme, chimed in with a comment saying:
Techmeme’s algorithm is so complicated at this point that I could write a long article on various ways linking can help you and/or hurt you, ranking-wise. Instead, I’ll just make the following vague claim: linking to new news in a way that doesn’t appear to be an act of quid pro quo should never hurt you.
…which is the same sort of cryptic non-answer that we have come to expect from him.
My guess would have been that there seems to be a penalty for linking to other blogs (IE blogs that don’t link get preferential treatment) for two reasons- 1) big sites don’t link out to smaller ones yet bigger sites or higher traffic sites are the ones that often the ones that get the headlines in Techmeme. Just going down the list right now the main headlines are from NY times, Mashable, Eweek, Next-Gen.biz, Macworld, Cnet, WSJ… 2) If your blog breaks the story there is really no one to link to, changes are they are all linking to you. This would mean you get the big link from Techmeme not only because you are first but because you have a lot of incoming links and people quoting your headline etc. This theory seems to have been debunked by Scoble in the comments of his post though.
I do know that it takes more than linking to Techmeme a few times to get them to link to you. The Memeorandum Bot comes to my blog pretty often (according to Feedburner) but I rarely get links. I am probably too late to comment of stories most of the time, but even when I do get in early or create the news I tend not to get the link. It’s all interesting, especially in an academic “how the web works” sort of way but not something that is going to make or break a site.
