I was reviewing the stats relating to my last post about the performance of Social Media Releases earlier today. Apart from being very pleasantly surprised to have nearly reached 1000 unique views (making this easily my most visited post to date!) I also found that the number one source of traffic was “direct”.
This means traffic that has been referred to the urls directly and not from another online source. Now unless anyone is able to remember a url with 141 characters, which I am guessing is unlikely, then I assume all of these visits must have come from people visiting the post from emails that include a hyperlink.
The breakdown of unique visits and their sources is as follows:
| Source |
||
| Direct/mail | 481 | 49% |
| 284 | 29% | |
| Blogs | 83 | 8% |
| Search | 68 | 7% |
| 42 | 4% | |
| Other | 23 | 2% |
| 981 | 100% |
Twitter also counts for almost a third of the visits, which isn’t surprising given that the post has been tweeted 128 times to date (thank you to everyone who RT’d).
But the fact that almost half of all visits were direct demonstrates the power that email still has as a tool for sharing information with your community. Glad we didn’t forget to mention it in our video (2:25) on how news spreads online 
I think you missed one important factor–that click throughs from Twitter desktop apps *also* count as direct. I’ve seen a few different stats, including one saying that 90% of active Twitter users use a desktop app, I think that is probably where all the direct traffic came from, if it was retweeted 128 times.
Hi Kelly Thanks for stopping by. I wasn’t aware of that but now you say it, it makes sense. But that begs a different question. If a large proportion of the directs are from Twitter apps and 90% of active Twitter users use an app then how come so many visits still came from Twitter.com as well? Does this mean the post was proportionally more interesting to less active users? Or am I being dim?
Well I’m not sure 90% is the most accurate number, just one I’ve seen (of course I’m *trying* to find stats now and can’t at all!)
Also, I use Nambu (Twitter app) but I often switch between it and the web interface, 1. because it crashes all the time, and 2. I sort of like the web one sometimes… Retro! So maybe others are in the same boat?
What you’d need is a Twitter analysis tool that could tell you # of clicks from all of Twitter and then you could cross reference. (Totally not trying to pitch my company, but we do have such a tool and it’s very cool! Been playing around with it all week)
I’d run a little test for you, but looks like the post is more than 10 days old, and unfortunately Twitter adjusted the API (& search.twitter.com) to only archive for 10 days.
So if you run a similar test in the future let me know and we can see how much of the direct traffic is from Twitter and how much from other sources.