Polemicists occasionally recruit “facts” in support of their cause, and I find it restful to examine those off in a quiet corner while the screeching continues.
A case in point is the ongoing discussion about Daily Kos, and the supposed “fact”, in a diary linked to below that web traffic there is not growing and is less than that enjoyed by Freerepublic.
For those who don’t want to read the extended box and supporting data, here’s the “shorter” version.
Web traffic comparisons based, as these are, on data from Alexa are suspect at best. Based on number of participants, not visitors, Daily Kos grew from the beginning of the year to just about the time of the pie fight.
You may now resume your regularly scheduled jousting.
Still here? Ok, returning to the diary Why DailyKos isn’t worth …, eodell says the following.
I looked at the cited link, saw the numbers matched what was said, but I didn’t believe them. I didn’t believe them because I’ve been tracking participation at Daily Kos since last fall (the day of the election if you must know), and I’ve seen the growth in participation there this year, so large I’ve proposed it as the root of the “problem” that so many people seem to be able to find there currently.
But back to web traffic. I have studiously ignored that subject until now, so if what I’m about to tell you is not news, my apologies, but it was new to me.
Plainly put, the numbers that eodell relied on aren’t valid. He gets his numbers from Alexa. Alexa gets their numbers from the Alexa Toolbar, available from Alexa for Internet Explorer on PCs only. The Alexa Toolbar installed base has been characterised as “a very limited audience sample, strongly slanted to represent … online marketing specialists, web site owners, webmasters”. Given the recent increase in other browsers (currently 48% at Daily Kos are non IE according to SiteMeter) and the limited and skewed installed base of Alexa data providers, Alexa simply can’t provide good numbers for either the site to site comparison or site growth over time.
Eodell could have said that the DK traffic coming from people using internet explorer equipped with the Alexa toolbar has been stable and isn’t growing, and is 1/2 to 1/3 of that same sort of traffic to FR. Not such a strong statement, is it?
For anyone interested in learning more about comparing web “status” of various sites, I recommend this article: PACmeter – Popularity, Authority, Credibility Online: How To Measure Them?, wherein you can find this prononouncement with respect to Alexa measurements: “These lopsided and absurdly inaccurate traffic rankings in themselves are enough to warrant a chuckle”.
Other ways to compare web sites:
Number of inbound google links: DK: 64700, FR: 23100, BMT: 316
Google PageRank: DK: 7, FR: 7, BMT: 5
Technorati link ranking: DK: 3, FR >100, BMT >100 (Instapundit: 2 by the way)
I hope for Boomantribune continued growth to a good and sustainable size. For Daily Kos, perseverance through troubled times. And for everyone, data you understand and believe in.
It’s interesting and important totally apart from the pie thing.
I was hoping it would be taken that way. Hey, I’m from Nashville.
If you haven’t already seen it you’ll probably enjoy this WSJ article
[http://online.wsj.com/public/article_print/0,,SB111685593903640572,00.html
Measuring the Impact of Blogs
Requires More Than Counting
Hey jotter. Glad to see you over here as well.
Thanks for the info.
All you pioneers make it easy for us latecomers!
There are those of us who use Firefox + adblock to block sitemeter, and hence, not only don’t view ads but aren’t tracked either. Not say’n I’m one of ’em. *cough!* –M
I am not fighting for blogs, but rather for the issues.
Yes that figure looked suspicious to me because I remember reading that no one topped Daily Kos for traffic. Thanks for explaining that.
so large I’ve proposed it as the root of the “problem” that so many people seem to be able to find there currently.
What figure would you propose for the optimum number of active participants on a political blog?
I think that membership that is inactive should be deleted and their numbers given to new registrants.
That would better reflect the actual population of a blog.
and even more important, not fighting within blogs.
I think the right size has to be determined by trial and error. For Scoop blogs like DK, BMT, and ET, I would look primarily at the number of users either posting diaries or recommending diaries over one week, and start thinking about making changes when that number goes above 3500. For reference, last week at Daily Kos, that number was 4422. At the beginning of the year, that number was 2154.
I have to disagree with you about reusing user id. We’re not going to run out of numbers so everyone should get a new one. It is reasonable, though, to keep track of and publicize the number of active participants.
This week on daily kos there was a good opportunity to look at that, around the 2008 presidential preference poll. On Monday there were 284 diarists, joined by recommenders to total 1638, then joined by commentators to total 2267, and finally joined by poll voters to total >10,000 (exact numbers not meaningful because the poll didn’t stop accepting votes at the same time I stopped counting diaries).