Valleywag
Back during the last boom, boastful internet sites used to refer to the "hits" shown by their server logs, as if that was a measure of popularity. They glossed over the fact that every image on a page registered in the raw statistics, and so every single page viewed could easily add ten times as many hits. In 2006, there's a new fiction: that every feed downloaded, even if the content is never seen by a human being, counts as web traffic. And Federated Media, an ad network founded by one of the internet industry's most luminous luminaries and backed by the New York Times, among others, is guilty of it. For the proof:
Federated Media, which competes with Gawker Media in business model, if not directly, is described, most recently in a release by one of their partners, thus: John Battelle's author-driven network reaches tens of millions of unique readers with influential blogs covering technology, parenting, media & entertainment, sports and automotive.
At best that's just a misunderstanding, at worst it's self-delusion, and most likely it's a lazy boast they expect nobody can be bothered to correct. In October 2006, according to Comscore's Media Metrix panel data, Federated Media — which reps high-profile properties such as Boing Boing and Digg, and also included humor site Fark under its umbrella at the time — had an estimated 5.2m unique visitors in the United States. Hardly tens of millions. even if one makes allowance for international audience, and a few hard-core feed addicts. Here's the Comscore screenshot, with comparisons between Federated Media, Gawker Media, and AOL's Weblogs, Inc. blog group, for disclosure's sake.
Before anyone quibbles, Comscore's sample of over 60,000 people does include Firefox users, and it does reflect those scanners of feeds who click through to view the full article on a web page. Comscore Media Metrix is the statistical reference of advertising buyers. It's got its flaws, particularly in the measurement of smaller sites, but it's way more balanced than web services such as Alexa, and more trustworthy than claims based on unverified internal stats.
So, why should anyone care? Well, this is the internet's own version of the circulation scandals that have holed print industry. Advertisers are finally coming round to the idea that the web is more measurable than traditional media. It's a major change. Advertising spending is shifting online; ad networks such as Federated Media, which reps some excellent properties, can succeed just fine without exaggerating their audience; there's no need to fluff the numbers; it's just tacky.
Comments
Is this post connected to your old feed suddenly changing from full-content to prematurely truncated posts?
http://feeds.gawker.com/valleywag/full is no longer complete...
http://valleywag.com/index.xml is equally bereft
Does Federated make any money? Where are their ads on Digg, I can't spot them. Who are their big advertisers?
Irrespective of any of the page views number claims etc. etc. I will say that as a blogger in the FM Network that they do amazing work for me. I'm certainly nowhere near the high profile of blogs like Boing Boing or sites like digg, but they consistently fill my advertising space with quality ads at some of the highest CPMs in the industry.
As an author I literally do nothing but say yes or no to quality advertisers that they bring to me and collect checks. I don't even have to place the ads in my template. Their team does everything.
The website they use to track my ads works very very well and they provide tons of support and other tools include good author forums and feedback on how to make my blog better.
John Battelle's done a great job with the network. I've been very happy working with them over the course of the past 6 months or so.
From Chas Edwards, vp sales, Federated Media: I just read your post on FM's inflated numbers claims, http://valleywag.com/tech/cold-stats/john-battelles-inflat...
We're in frequent dialog with the Comscore gang (and have recently become paying customers!) in an effort to find a better alternative to server logs when we talk about the aggregate reach across our family of sites. But we're still not satisfied with the data. In a chat with Gian and Magid at Comscore earlier this month, even they admitted that accurate measurement of small sites (or even large sites like Digg and Boing Boing, which skew towards tech savvy young men) is a code that hasn't yet been cracked by anyone. I'm sure Comscore will eventually get this figured out, and we're doing all we can to help them.
It will be a great day for all of us when the audit firms get this right -- and capture the reality that upscale, web savvy consumers are gravitating toward Gawker, Gizmodo, Digg, Boing Boing, Dooce and others, instead of the traditional favorites from a decade ago!
If it were a few weeks ago, the headline might have ready "Battelle farks numbers...."
Yeah, that was definitely the low hanging fruit.
I gave them my numbers a while ago straight from my own stats, and they printed them on their page covering Google Blogoscoped -- not much, not less, no mystery:
http://www.federatedmedia.net/authors/outercourt
Now when someone says "unique readers" it starts to become meaningless, because you should either say "page views" or "unique visits", and when you say unique visits then you might wanna add how you count 'em (e.g. "same IP within 30 minutes range").
I sit on my stats, and it's absolutely impossible for me to find out how many different people I'm reaching with my writing. Some of these blog posts are copied on other networks (I use a Creative Commons license), the RSS reader completely hides the number of views, someone looking at the frontpage may read 1 post or 20 or none at all and move on to the blog's forum, the same person will have a dynamic IP and thus not be recognized as such, and so on and so forth.
BoingBoing's awstats log (http://www.boingboing.net/stats/) shows 2.3MM uniques and their Feedburner feed shows 3.2MM readers.
I tend to believe both these numbers. So the claim really comes down to how many people are using Digg and I think Battelle got caught up in that nonsense. I'd imagine the conversation at Digg Central might have gone something like this:
"Kevin? Here boy, sit down. Have a biscuit.
Here's the good news: your personal share of the
Federated Media money -- even after we deduct the
beer expenses and the cost of having the corporate
furniture de-loused after the Micki Krimmel incident --
still averages $21,000 a month.
The bad news? Well, the statistical dickwaving required
to close those ad deals pretty much means you have no
credibility and can never sell the company or involve
yourself in any media-related venture, ever.
Aw, little fella, don't cry. Have another pumpkin beer."
Whoa! awstats numbers are usually garbage. The tool does a lousy job of figuring out that multiple images load in the same pageview.
To get the real number, something that SiteMeter would give you, you have to divide awstats by 5-10 or so...
Comscore's sample of over 60,000 people
anyone who buys into such a flawed methodology gets exactly the level of accuracy they pay for. comscore sucks.
I beg to disagree Sample My Taint, I've looked at everything out there and Comscore is the best there is.
And Nick has his facts wrong.
Comscore has Federated's worldwide uniques at almost 11mm in October.
Nick's data in the post is US audience only.
Fred
ScalaWag wrote: "Whoa! awstats numbers are usually garbage. The tool does a lousy job of figuring out that multiple images load in the same pageview."
Not sure what that has to do with counting IP addresses (the number I gave that most any dopey stats program can handle) but I'm sure you absorbed that tidbit of useless knowledge on a forum somewhere and have been dying for a chance to spew it out.
If you want to add value to the conversation, here are the dreaded awstats and, as a bonus, a little vintage Battelle talking about them. I look forward to further apples to orange analysis from you.
http://www.boingboing.net/stats/
http://www.boingboing.net/2005/01/04/boingboing_traffic_s....
Start a discussion:
Login with your username and password below. Or comment on this post via email.
Forgot your username or password? New User?