Reading this story by the Deal about the gloomy outlook of Friendster, I came across a chart showing the unique visitors of US social networking sites (provided by Comscore, as of March 2006).
Quite stunning to see that 23% of US Internet users are visiting MySpace (1 in 4).
It would also be very interesting to see the percentage of overlap among visitors of these services.
David Hornik (sitting in fronf on me) tells that combining visitors to Six Apart’s LiveJounal and TypePad would make them rank pretty high on that chart (granted these are blogging platform more than social networking sites).
Aliens invade MySpace!!
Ok, who's keeping track of ways not to report web traffic numbers.......
I'm just going to take a stab and guess that MySpace didn't have 41,889 *Million* Unique Visitors last month. :-)
Ok, so being an Excel junkie, I know they had their number format stuck in European mode, but, with this going to a US audience I think its fair to ask for a US standard number format....
Even then, I have a problem believing these numbers, even in European format. Basically they are claiming 41.9 million US Unique visitors. Is that even probable?
I thought there were only about 70 million households with Internet access in the US. That would put them at nearly 60% of all US households with Internet access ........ hmmmm.....
This then suggests that they are double counting the same people, a) same person different computer, b) same person, same computer, different cookie, c) same person, same computer, same cookie, different user info...d) or have they maybe figured out how to tell the difference between people using the same computer, or even more unlikely, but very cool if true, e) they do have 60% of the US market visiting every month?
That said, if they are using the same basic methodology as the rest of us, what are we all going to say when they start reporting that they have more unique US visitors than there are unique people in the US?
-bruce
Posted by: Bruce Boston | May 04, 2006 at 09:26 PM
They use spyware, just look up netsetter and marketscore. They have 2 million installs where they record browser habbits. At any rate the numbers they give publically are worthless. If you want real stats their fees start at $40,000
Posted by: markus | May 05, 2006 at 12:07 AM
hmm, well we are definitely top 10, but i guess media metrix are ignoring The Buzz...
Posted by: marc | May 05, 2006 at 10:01 AM
Ok, so they are using a basic panel-sampling method based on spyware results....
So, I'm guessing that they do a number of calculations here. 1) to find the total US Internet Population, and then 2) at least one more to figure out what percentage is doing X, like visit MySpace.com.
In this case, they have figured that there are 171,421,000 Americans on the Web, and that exactly 24.4363% of them surfed MySpace that month. Is that +/- .0003% accuracy?
But this also begs the question, doesn't this suggest that their numbers will skew towards populations that don't know how to remove spyware?
Also, what does 2MM installs mean? Any one know how many of those installs are active today, or over what period they have been doing this?
While this may not be the place for a discussion on Theoretical Web Analytics, I'm wondering how much investing is being done on these types of stats?
Does the likelihood that your site's community will download and keep spyware on their machine increase the likelihood that your site will get funded? Fetch a higher exit price? If a company like Facebook really want to catch up with MySpace at all cost, why wouldn't they start encouraging their their community to go download the spyware? Who's doing this already today?
Sorry, the economist in me is wondering what the incentive trail is here, and what it might mean to the flow of investments.....
-bruce
Posted by: Bruce Boston | May 05, 2006 at 12:32 PM
Nielson and comscore use spyware and their numbers are often out by a factor of 10 when comparing sites. From looking at google searches it looks like most of these installs were done via file sharing apps such as kazza edonkey etc
Hitwise.com works by getting a nightly import from ISP's who hand over all your surfing data.
2 months ago, the 6th most visited dating site was supposidly a 1 page site that collects phone numbers. In the end the data is only worth what people think its worth.
Posted by: Markus | May 05, 2006 at 10:15 PM
Its embarrasing that neilson ignores so many sites when it ranks things. Plentyoffish.com and faceparty.com. It shows how large corporations don't get it. They only look to the big headlines. Again I want to second Markus, hitwise (a portfolio company of my firm) does a much better job. The data is all skewed anyways, you should rank free social networks by monetization. That should give you a better picture of what really matters. Facebook is crushing Myspace then and for good reason. The key is tying offline interactions with the online world. They are interrelated and very important things.
Posted by: Bradley | May 06, 2006 at 07:13 PM
I would like to note that hitwise's data is collected Anonymously. So there are no privacy concerns. The information provided to subscribers never reveals any personal information because hitwise does not have nor is interested in any of that data. It does do a great job collecting statistics even though a few outliers make their way to the top.
Posted by: Bradley | May 06, 2006 at 07:29 PM
The killer social networking which is going to blow away competition is here live and kicking... www.aimpages.com is the killer app of next gen with its ModuleT... users can build their own modules and more than that they can link application from other social networking and cool applications... Innovation at its best... watch of other social networks... start losing your page views....
Posted by: Web2.0Guru | May 11, 2006 at 09:55 AM
The next application that is worth watching in this space is www.blinklife.com
Posted by: Mike | May 14, 2006 at 08:53 AM