Where My Traffic Comes From
Every once in a while I like to show some analytics about this blog.
I get the sense that the sources of traffic to this blog have been changing recently. Let’s start by comparing the high level sources of traffic to this blog in the first quarter of 2007 vs the first quarter of 2008:
First Quarter 2007
First Quarter 2008
The total page views didn’t change that much from the first quarter of last year to this quarter, growing from 400,000 page views to 450,000 page views. But the sources of traffic changed for sure. Search traffic was down from 50% to 40%, and direct and referring traffic both grew by 5%, and referring sites now drive as much traffic to this blog as search. That’s a big deal. And I expect that trend will continue.
So what are the top referring sites and how did that change in the past year?
Q1 2007
Q1 2008
As you can see, the sources have changed a lot. First, Google in this context (at least the Google that is number one in both charts) is Google Reader – the king of RSS readers (at least for my readers). The other RSS readers on the lists include MyYahoo, which went from number 2 to number 9 over the past year, Bloglines, which went from number 3 to number 7 in the past year, and netvibes, which was number 8 in the first quarter of 2007 and is no longer in the top ten.
This post should be called "the rise of the smart aggregators" because the new big sources of traffic to this blog are techmeme, reddit, twitter, and delicious. Both techmeme and delicious were on the list in 2007, but techmeme jumped from number four to number two and tripled the number of visits it generated. Delicious stayed at number six but grew the number of visits it generated. Reddit is the real eye-opener for me. It wasn’t in the top 10 in 2007 and is currently the number four referring site. Likewise, Twitter was not on the list in 2007 and now is the number five referring site. And check out Hacker News (news.ycombinator.com), it brought this blog more traffic than my.yahoo this quarter. Thanks Paul and everyone at ycombinator.
Some things don’t change. Stumbleupon remains an important source of traffic and there’s always some big source of traffic that is probably a link to a single post. In Q1 2007 it was a cnet link to a post I wrote about not being able to buy an iSight. This quarter is a google images link that I can’t track through to the post.
There’s an important new "smart aggregator" out there, FriendFeed, that doesn’t show up in the Q1 numbers, but this month to date it is the number seven source of traffic to this blog.
March 2007 (to date)
You’ll also note that in March, the ycombinator connection is even more pronounced with reddit ahead of techmeme and hacker news in the number five slot. It’s also nice to see my tumblog being used as a way to follow what’s happening on this blog.
Finally, I should say something about the direct visits. With a blog URL like avc.blogs.com, I’ve always felt that direct visits are most likely driven by browser bookmarks. And it’s also true that a good deal of the search traffic is a proxy for a direct visit.
Google sent almost 120,000 visits to this blog in the first quarter from almost 50,000 search terms, so it’s a very long tail. But the top ten search terms are shown below:
I figure that searches for a vc, fred wilson, avc, and fred wilson blog are basically "search bookmarks" and so 9,500 visits or 3% of my search traffic is really direct traffic.
So what is the big takeaway from all of this? First, that the two big non-direct sources of traffic to this blog are search and aggregators. That’s always been the case. But the second, and possibly more important point is that the aggregator market is changing from rss readers like Google Reader, Bloglines, Newsgator, etc to smart aggregators like techmeme, reddit, hacker news, friendfeed, twitter, delicious, and stumbleupon.
That makes perfect sense to me. I’ve never been able to use a RSS reader, as hard as I’ve tried. But I use all of these smart aggregators every day.
The other thing that is interesting to me is not one traditional content site/blog is in the top 10. Not techcrunch, valleywag, venturebeat, silicon alley insider, or other blogs that occasionally link to this blog. Techcrunch is number 13, Silicon Alley Insider is number 18, Valleywag is 20.
And just to show that this is not all business, Howard Linzon is 21, Brad Feld is 22, and my wife, The Gotham Gal, is number 26 (she sent me almost 1000 visits this quarter, more than read/write web, seeking alpha, and the new york times).
Anyway, that’s all for now. I hope you enjoyed a sneak peak behind the curtain into google analytics.
Comments (Archived):
This is really interesting and incredibly helpful just for context. In the last few days I have been very curious about these questions as I just started my blog and was curious how some of my ratios compared to others.One other statistic which I am curious about is your RSS conversion rate. I started my blog in january and over the first 3 months I have about 380 subscribers and I have about 36000 unique visitors over that period of time. Both in the aggregate, and month to month my conversion rate to RSS is about 1%. I am curious if yours is similar now, or has been in the past. Is that a statistic that anyone cares about or tracks? Perhaps I am over estimating the significance of RSS readership but it seems to me that understanding what typical conversions are and what effects it could be really useful.
Feedburner says I have approx 120k feed subs but I think most of them are either bogus or inactive. Feed subs is sort of irrelevant. I care much more about the “daily reach” of my feedThe ‘daily reach’ according to feedburner is around 4000 unique viewers of my feed. That is about equal to the number of unique visitors each day to this blogSo what that means is half the audience reads this blog here and half reads it somewhere elseIf you don’t use feedburner, I highly recommend itFred
Thanks. I do use feedburner, and it is indeed helpful. I guess my blog isnew enough that the daily reach and the subscriber numbers are pretty closeand at this point I dont think I have almost any bogus subs. Currently thenumber of daily readers is much higher than my subscribers or reach. I guessover time it sounds like that dynamic will change. What I was trying tofigure out is are there metrics for how many “conversions” one can typicallyexpect out of a batch of uniques that will become subscribers. My *tiny*sample suggests in my case about 1%, which I am pretty happy with. But I wasjust curious if that was good or bad my comparison. It probably depends onwhere you are in the lifecycle of your blog too I would guess.
Exactly
Once again, Google shows they are the biggest game in town.
i knew rss readers were over-rated… 🙂
Everything is overrated unless it’s unknown
i was one of those folks completely confused by rss. the words involved were too long (aggregator) and completely foreign.Now Reddit I love. It’s (unfortunately) what I am reading while all my friend are on Techmeme. I remember when Reddit was bought by Conde Nast a couple of ears ago for a song. As a long time user of it it seems they didn’t touch it at all (except to slap some unobtrusive ads on it).But back to th topic at hand. I think its cool that you let people know this stuff.It’s interesting.I am currently working with a fairly large site and we were discussing their stats. I advised them to open their stats up to everyone.They freaked and talked about stuff like competition knowing, etc. My answer was the standard “So?”.They didn’t have a reply. From now on I will use the line I heard waiting for the train today. “Don’t be scared. Be scary.”I know you didn’t post this stuff to scare people but being open is surely more interesting. A great read.
That’s a great lineI can’t think of one time in the past fice years where I’ve regretted being openFred
That is a great line and it’s true, but you’re not making (the bulk of) your money via advertising on your blog.I had some e-mail correspondence recently with some folks about NewTeeVee.Com’s page view traffic because I could see it via Quantcast. The result of that seems to be that Quantcast now only has that level of detail for all of the GigaOm empire rolled up (as opposed to pageview detail on the individual sites). So it’s still “open”, but not as granular.The interesting info is actually in the details sometimes.. But, I don’t blame GigaOm at ALL for making that change. Nor would I expect you to post your pageview detail or pages per visit simply because I’m curious, but it wouldn’t break my heart if you “Quantified” your blog leaving it all open to anyone who cares to look! They also have a way to report feed stats as well, though I’ve never played around with it.
Quantcast’s data sucks. It makes no sense to me. I prefer google analytics, sitemeter, feedburner, even alexa if you can believe it
Hmmm. I think it depends on whether you’re “quantified” or not. If you’re not, (which you’re not) and they are relying on panel data, yeah, at least for the smaller sites, the panel data does seem to suck (but I imagine that’s true on almost any panel).But if you’re Quantified, at least on a visits, unique, and pageview basis, Quantcast tracks almost exactly with GA. It tracks at least as closely as feedburner does. I don’t love the presentation of it’s demographic data but for measuring the traffic itself, it’s very good.I’ll grant there are some presentation/UI issues, but if you look at daily traffic (instead of 30 day or 7 day trends), it’s pretty spot-on.To compare it to Alexa for a site that actually is quantified is insulting.It’s definitely NOT better than GA for the person running the site (but it’s not trying to be that) because it doesn’t list all the referral, keyword, and wealth of reports available on GA, but in terms of making your traffic stats totally open, if that’s really what you’re of a mind to do, it’s great.
I had the quantcast tag on my page for about a year and I took it off maybesix months ago because the data made no sense to me. Maybe I was doingsomething wrong. I have a link to my sitemeter stats on the left rail of myblog. That seems to work fine for me and others. all of gawker media statsare shown that way on nick denton’s bloghttp://www.nickdenton.org/0…
Good Article. You get one thing wrong though imo. All the traffic sent to you through GoogleReader does not only come from your own subscribers but also from people subscribed to tc, rww and the likes and clicking on a link to you in one of their articles. So, to sum this up, you don’t really know and will never know how much traffic these techblogs with an rss-affine audience really do send you. 🙂
Since you put the entire post in the RSS feed, it probably means that most of the RSS audience never clicks over. This is as opposed to techmeme, del.icio.us, reddit where only a portion of the post or only the link is there. I read your blog inside Bloglines and almost never click over.This is not to say you should put partial posts in the feeds. Techdirt has a good explanation why full posts actually increase pageviews – http://www.techdirt.com/art…There are still a few blogs out there (chartreuse) that don’t put full posts in their feeds. I’ve stopped reading them as they don’t fit into my consumption experience which is increasingly through the iPhone (Bloglines has a great iPhone site)
I agree. Full posts in feeds is the way to go. You can monetize via feedburnerFred
Excellent pointThanksFred
Fred,If you could post your stats breakdown from FeedBurner, that would be interesting to see and compare.
I still use NetNewsWire as my aggregator. I like it because it’s a separate application like an iTunes. However, still waiting for the Joost online browser-that’s a program I prefer not to deploy a separate app.
Fred,What were your total unique visits? You mention page views but how about sharing the unqiue number? I’ll share mine if you share yours.Thanksjohn
Around 75k per month last time I checked plus whatever the uniques are that read my feed and don’t come to this page (like dan who left a comment earlier)Fred
i just checked google and it was 83k for the past 30 days
Thanks Fred. I’m a man of my word :-)furrier.org blog did 18k uniques for the qtr around 6.5k per month … i have no idea how many subscribers are hitting the feed.
it’s amazing how fast FriendFeed has grown as a referrer, almost catching up Twitter (and I suspect mostly at the expense of Twitter)
That’s not what my data shows
that’s good new because it means FriendFeed is (I guess) introducing a new audience to the social networking world.digging into my (much, much) smaller pool of data actually bears that out… I have FriendFeed followers who are not Twitter followers… they came from Facebook.Kinda comforting to see it’s NOT a zero sum game as I first suspected it might be
Google Analytics= Nerd Porn (I mean this in the best possible way!)
So true. And office safe too
Have you read “The Open Kamona’ by Seymour Hair 🙂 . Oy.
Love to see your top 10 from “network location”.
I am interested to see the relationship between Feedburner subscribers and Twitter Followers for any correlation as to early adoption and success. For instance, I liked reading Gaping Void dude for a time, until he twittered and once I saw how he twittered I lost interwst in everything he did. I think some people are meant to twitter, some blog, some both and some just to read :)To start this conversation off I have 3400 feedburner subscribers and 530 or so twitter followers for a ratio of 6 ish to 1. I like twittering better for finance and blogging now for other stuff.
HowardThe problem with feedburner subs is they get stale. I have 115,000 feedburner subs. And yet when my subs were 15,000 the daily reach of my feed was around 2,000. Now its about 4,000. So my feed readers have doubled and yet my subs have gone up 8xThe feed sub number is like the facebook installed apps number. Basically uselessFred
a concrete example of how much deterioration my.yahoo has experienced over the course of only one year at the expense of iggogle et al…
The vast majority of my traffic comes from festish sex porn sites that cater to midgets, tri-sexuals and fixations with small furry mammals.I don’t understand this at all.My site is geared towards conservative Republican values.
Best comment of the day!
Fred – great post. I have a theory about why blogs of similar interest like techcrunch do not generate that much traffic to your blog. The explanation might lie in the similarity.Since I am interested in Technology, Entrepreneurship, and VC, I am subscribed to most of the usual suspects like GigaOm, TC, your blog, etc, and go through them via Google reader. If I find a link to one of your posts on, say TC – there are a couple of possibilities why I might not click that link:1. I have already read the post2. I will get to your blog and read the post as I make my way through the subscriptionsJust a thought – perhaps this is deflating the number of referrals from sites that appeal to the same target audience.