A week ago Fred Wilson dove into his blog’s stats to answer this question: Does this blog get more traffic from Google or Twitter? On the surface, it appears that Google drove more — nearly 15% of his visits versus Twitter’s 8%. But over the past year he’s also seen an unexpected jump in his “Direct Traffic,” visitors who start a browser session at his site rather than linking in from another website.
“If just 7% of the 17% increase in direct traffic is a result of twitter links that are not being counted as twitter [because many Twitter users use desktop applications like TweetDeck that wouldn't show up as a referral site], then it is true that this blog gets more traffic from twitter than google.”
It’s a believable scenario given that @fredwilson has more than 27,000 engaged Twitter followers. Results may vary, though, since we more average Twitter publishers have only 126 followers. For PC Mag (with fewer than 7000 followers), Twitter is a much less significant traffic source, less than one percent of visits to Google’s 4%. But like Fred Wilson’s AVC, PC Mag’s top source of traffic isn’t Google, it’s a sharing site — Digg drives 81% of the visitors for some articles.
And as Bit.ly’s Andrew Kortina explains, most Direct Traffic to most sites isn’t really direct traffic from loyal readers that bookmark your site or type its URL into a browser. It’s traffic generated by readers sharing story links with her or her friends and followers.
“The only reason hits to pages other than your homepage show up under Direct Traffic is because they do not have an HTTP Referer Header. Typically, in your browser, when you click a link on Page X to Page Y, your browser sends a bit of data along with the request to fetch Page Y from wherever it’s being server that says, ‘by the way, this request was referred by Page X.’ Desktop applications — email clients like Outlook, for example — do not send this referrer data, so clicks derived from these apps get grouped under Direct Traffic….
“In fact, very little Direct Traffic is actually direct. A better name for Direct Traffic would be Shared Link Traffic — think about the apps that are likely to generate this type of traffic: email clients, twitter clients, desktop RSS readers, facebook clients. These are all essentially sharing tools.”
I followed Kortina’s instructions to calculate ChasNote’s Shared Link Traffic. Here are my traffic sources according to Google Analytics:
The message I take away from this overview is: Great, I’m pleasing 17% of my audience enough that they come back on their own. It’s nice that 46% of my traffic comes from editorial pick-ups by other sites (Referral Traffic), mostly blogs somewhere out there in the Long Tail, but I can’t do much to influence that. So I’m going to spend my time and resources getting more from Google. I better optimize my URLs, page layouts and story headlines for search, and if I ever invest in online marketing, the first call I’ll make is to Google.
When I isolate Shared Link Traffic — removing so-called Direct Traffic that lands on a story page (URLs too long for anyone to type into a browser and too static for anyone to bookmark) as well as Referral Traffic that’s coming from sharing platforms like Digg, Twitter and Facebook — I come to a different conclusion.
Forty-eight percent of my traffic is Shared Link Traffic. In my case it’s four sources — Digg, Twitter, TechMeme and Facebook — plus (I’m guessing) email sharing. This changes my strategy. First, I’m not doing as well as thought with loyal readers: They represent 8% instead of 17% of my audience. Second, I’m going to change my optimization priorities. Since Shared Link Traffic is 48% (to Google’s 37%), I should focus on “social media optimization” — making it easier to share my stories — before I invest in search engine optimization. And when that request for the ChasNote marketing budget clears Finance, I’m going to call Twitter, Facebook and myself here at Digg before I call Google.