Twitter traffic via FeedBurner Socialize now represents majority of time on this blog

Recently, over at Google we've done two interesting things that allow you to get a little more information out of referral traffic to your site. FeedBurner can now automatically tag your feed to make it compatible with Google Analytics, and we now also allow you to automatically distribute links to Twitter. The combination of these two services is even a little smarter than that, because when FeedBurner knows it's sending the feed out via Twitter, it marks it as such, so you can actually tell the difference between traffic that is organically distributed via Twitter versus traffic that originated from the feed.

In the case of this blog, and like most weight loss pitches, these results may not be typical - traffic originating from the feed and coming from Twitter seems to be of better "quality" than other traffic, as is now resulting in a majority of time spent by users on the site.  

From a standpoint of just comparing the medium from which this time on the site comes, Twitter is the clear leader. In this case, the difference between "Twitter" and "feed" is simply those clicks that came from the socialize service from the feed, and clicks from feed readers. The big caveat here though, is that I publish full content feeds, and most of those views happen in Google Reader, so there is little incentive from for the user to come here and spend time on this site.  As we've noted many many times before, users of Google Reader and other feed readers tend to be a separate audience that stays in their RSS reader when reading content.  Many won't even subscribe to partial feeds.  The combined aggregate view of time in Reader plus this site would be much more interesting and something we'll have to work on providing.

From a Source point of view, you can see that most of this time, including the traffic from Twitter originated from the FeedBurner feed.   Look at the big yellow PacMan in the picture below.

Breaking this down a little more, you can see even more that the Twitter traffic itself is divided between the twitter.com site  (Twitter) and other non-web Twitter clients ("not-set"), which I've confirmed via other analytics on this site are mostly iPhone and Android Twitter clients.  According to PercentMobile, about 7% of my traffic is mobile, and I think that's all referrals from Mobile Twitter applications such as TwitDroid, TweetDeck and Tweetie.

We're working on ways of being able to display this breakdown as well so it's not marked as "(not set)".

Finally, looking at both the Medium and Source together confirms that the majority of traffic on this site is coming from the Socialize service - that's the "twitter/feedburner" combo.

Again, these are just results for this particular site, and may or may not be meaningful for the larger ecosystem, but they definitely hint at some larger trends.  They are interesting nonetheless in showing that how that part of my audience interacts.  Clearly, Twitter users do click through on links and are somewhat more engaged than what is otherwise  "anonymous" traffic.  It's yet another audience that needs to be tracked and analyzed separately from search and other referrals, and separate again from those who consume content in feed readers.