Yahoo says Hitwise report is bollocks and should be discarded Full Story
((I fixed the spelling of bollocks)) -- Edit :)
It's bOllocks not bUllocks
Is there really any doubt?
even if Myspace do or will have more traffic it is comparing tyre kickers with shoppers.
I was thinking about posting a few thoughts about metrics and their limitations. Maybe I'll bump it higher on my list..
It's like saying that I have more searches than Google, but only commenting about (say) Google Base searches for Widget, found when analysing log files from a select few ISPs that have a tendency to use only my preferred search engine then extrapolating those numbers out as representative of the web searching world
It aint an outright lie but it is definately Bollocks!
What The City and Wall Street need is someone to educate them in search. Hmmmmm
to believe. They check in to myspace 4-5 times a day. My household visits Myspace on a 3 to 1 ratio vs. Y!. no prob.
there is an aweful lot of people loggin into their myspace accounts, we got a girl from the next office coming over to ours just so she can check her myspace a couple times a day... sadly, we wont go into how often i log in to check my comments as well... cough cough..
They didn't add together all of Yahoo's subdomains. Between finance.yahoo.com, movies.yahoo.com etc, MySpace isn't in the same league with Yahoo.
the spike makes sense. Its defacto calling card for kids/teens/young adults, I think they trade myspace info before phone #'s.
you get a bunch of teenagers to latch on to your goods, and you sell a gazillion records and make some of the old school bands album sales look like a joke...
Hitwise always breaks things down to subdomains, which makes their stats useless, but they aren't just useless in the way suggested by not counting all subdomains together. That would be skewed too, and even worse. For example the #1 referer to mail.yahoo.com is probably yahoo.com (I don't have Hitwise open in front of me now), while the a top five referer to yahoo.com would be mail.yahoo.com. A single person going through yahoo.com, mail.yahoo.com, my.yahoo.com and finance.yahoo.com could conceivably be counted a dozen times, depending on click path.
Hitwise is better than Alexa, but it is close to useless.
>> Hitwise is better than Alexa, but it is close to useless.
I don't think that's true. You do have to think a bit about what the data is actually saying, and be careful what questions you ask Hitwise, and how, but there's gold in them thar stats - for those that know how to extract it.
It's a tool, same as any other. If you learn how it works, it can produce good data
For all of Hitwise's problems it is still the best publically available data out there
Hitwise is also ranking Myspace above Yahoo and Google for video search.... number one being Youtube.com
But I shouldn't have said close to useless. That data is never correct, but it has its uses, most notably in comparing some thing to itself, like tracking a competitors traffic month to month. It's poor in comparing one competitor to another, and completely useless for some things, like comparing myspace to yahoo. That is just impossible due to them not offering any useful data for domains that have significant subdomains that interlink.
The nonsense Hitwise generates for my own domains is always good for a laugh, as is how they are fooled by popups, but I do spend the money for it, so it does have some value, if only to know what competitors are looking at.
Trolling through Matt Cutt's A Word About Metrics post I stumbled upon a link to Why Unique Visitors can be so different... by Matt Belkin of Omniture. It's a good read.
That is an excellent survey jimbeetle and an important topic. As long as we are on it...
If you were the type to seek opportunities to "capitalize on market inefficiences, profiting while helping to improve the market in the long term", audience metrics is ripe for the picking. The sampling methodologies are poorly understood among webmasters; only the most competitive webmasters can follow a technical sampling discussion. And so that means.... opportunity?
If there was ever a guy who was smart enough to know how that might be useful, and who for some reason has already demonstrated that he could if he wanted to (I'm not saying he does, mind you :-) it's Markus Frind.
*Active* Threadwatch Editors