Yesterday, at 2 o’clock in the morning, Pete Cashmore at Mashable tweets, Reading: “Why You Should Be Twittering More”. A bit surprising considering that I wrote that post back on February 5th, but not totally weird since he’s referenced in it, and it’s about the effect of a twitter run. Don’t get me wrong, I’m happy for the extra 2,500 visitors and 50 new twitter followers, but there was a more interesting part of the day – the data.

The “Twitter Effect” that hit my post yesterday served to crystallize some ideas I’d had for a while about using visitor data related to twitter traffic.
I use TweetDeck, and I know a ton of people access twitter without going to their twitter.com page. Using an application like TweetDeck means that when I follow a link, I get to that page with a blank referrer field. That makes the job of analyzing twitter generated traffic a little bit harder.
Back to my own data from yesterday. I know that the traffic to the post was as close to 100% twitter driven as is possible.
Knowing that lets me pull out specific data.
The post in question received about 2.725 visits yesterday. Of those 1,074 came from the twitter.com domain. There were also 1,353 that showed up with no referrer (aka direct/bookmarked).
There’s a chance that a handful of people happened by a post that during the previous month attracted a whopping 21 visits, but for this argument, let’s agree to ignore that.

That means that roughly 50% of the twitter traffic yesterday came in naked.
Does that mean we should all be multiplying our Twitter traffic x 2 to get a more realistic idea of it’s true impact?
I would love to hear about any other cases where there’s real data to lean on to make this argument or to refute it.