| This 40 message thread spans 2 pages: 40 (  2 ) > > || |
|Calculation of Alexa Rankings|
I don't understand how these are calculated
I subscribe to the Alexa database which displays rankings of sites that I visit, and the current ranking of my own site.
I don't understand how these rankings are calculated, which visitors they include, how often they are updated etec. etc.
Can anyone advise please?
|This might help a bit.|
I find the Alexa page explaining the background to the data collection short on detail!
It doesn't say how frequently the data on site ranking is updated.
It doesn't make clear if it gives a default ranking to sites not registered.
Some site appear to have a ranking of +14,000,000, I'm not sure if this means they haven't registered, or just that no-one has visited the site!
Many of us dismiss Alexa as a 5th-tier nuisance. Its spider is banned from sites (like WmW, I believe) because it burns bandwidth with little hope of producing return traffic. So the data they present has to be highly suspect, IMO.
Welcome to wmw, ChrisK.
I don't see how alexa can have ANY numbers on several of our sites. We've banned their bot for years.
BTW, I was being charitable when I said they were 5th-tier.
Alexa calculate thier ranking by the number of visits from people with the Alexa toolbar, nothing to do with the spider.
If you click more info, it tells you how many people visited with the Alexa tool bar in the previous 6 months.
Does anyone know which are the top ten visited sites according to Alexa?
read in the newsgroup searchengines that Google is fourth and Yahoo is sixth.
(Micorosft 18, MSN 21, Ebay 36, Amazon 38, CNN 61, Aol 74, Altavista 86, Hotmail 286,)
Cannot imaginge any sites doing much better than these - or is it just another quality flaw of Alexa's?
Dig around - there are charts that list them. I'm not sure where, but I've seen them before.
|http://www.venus.co.uk/~alec/BIG_search_engine_list.htm only lists the top search engines from Alexa data, but it is interesting all the same, I think they actually resell the complete list though.|
Having just taken a second look at it though, I think that list is well out of date!
carlwright, thanks for the list, is indeed a bit outdated.
I still think somethings wrong in Alexa's top 10 listing.
I'll bet a gin-tonic at the next London bar-conference for anyone proposing a site doing better than Google on position no.4.
Thank you for all your comments and help. The listing of search engines listed by Alexa ranking is very helpfull, even if it is somewhat out of date.
|(Micorosft 18, MSN 21, Ebay 36, Amazon 38, CNN 61, Aol 74, Altavista 86, Hotmail 286,) |
Cannot imaginge any sites doing much better than these - or is it just another quality flaw of Alexa's?
It's not so much a "flaw" as it is "excessive precision". Unlike a lot of ratings services, Alexa rates each fully-qualified domain name (FQDN) independently. In AOL's case, that means domains like [aol.com...] (Rank: 74), [search.aol.com...] (Rank: 765), and [members.aol.com...] (Rank: 28,742) all have separate rankings. (Alexa does "merge" some renamed domains. For example, it seems to know that "members.aol.com" and "home.aol.com" are the same site.)
So the "mystery #1 site" is either an FQDN we haven't guessed yet, or a dark horse candiate that doesn't split its traffic up as much as AOL and Yahoo do.
I actually came across #3 which was a Double Click ad popup but have been unable to replicate it.
We all know the #1 site is really webmasterworld.com
Google is now #3. Still no idea what #1 or 2 are.
What? No. That site has a Alexa traffic rank of 1,349,463.
sorry, I meant to say daum.net is the number one site on Alexa.
Where is my brain? :)
P.S. Google is up to #2 now.
Well yahoo korea (yahoo.co.kr) is #8 and yahoo japan (yahoo.co.jp) is #62. Hmm, methinks Korean users are way overweighted in Alexa.
Here's what I could find of the current top 25:
#1: daum.net - Korean email
#3: ad.doubleclick.net - Doubleclick ads
#6: damoim.net - Korean friend finder
#11: cgi.ebay.com - ebay's items
#12: Ar.atwola.com - AOL's pictures
Also, in my quest to find the top sites, I found tons and tons of non english sites and tons and tons of "non real" (secondary urls) sites in the top 250.
I woudl guess that the ones I couldn't find are a combination of subdomains from major sites and foreign sites.
|Hmm, methinks Korean users are way overweighted in Alexa. |
No great revelation. Even Alexa admits that in the second to last paragraph of http://client.alexa.com/destination_site/html/traffic_learn_more.html
"Many of us dismiss Alexa as a 5th-tier nuisance. Its spider is banned from sites (like WebmasterWorld..."
I wonder if that is a good idea?
I appears that Google is using SOME kind of information from Alexa - not sure what, but if you get the Google toolbar, it feeds back info to Alexa.
|I appears that Google is using SOME kind of information from Alexa - not sure what, but if you get the Google toolbar, it feeds back info to Alexa. |
Windsun, how do you know the google toolbar feeds back info to Alexa?
also, check Chris_r's remark at the bottom of this thread:
When did they start using Google's dB?
Was I sleeping again!
|brotherhood of LAN|
So lets say I took the info on there seriously
In your opinion(s), what do you think is an "average" margin for error for a typical site in there.........
and, since this margin of error will probably increase with smaller sites, give me a range :)
I am not aware of anything like that. If Google starts sharing information with Amazon that would be very big news. I am sure I would have heard about it.
According to Alexa reviews, my site "contains great Pamela Anderson pictures" and is "the bomb".
Quite an accomplishment for a site about SEO :)
I dont put much faith in the Alexa ranking numbers. I have a website that is not listed anywhere as it is still under construction. I have the alexa toolbar and as I am the only person visiting the site and can be the only one counted by Alexa, my site rank acording to them last I looked was 294,732. I guess that kinda blows their ranking methods if you ask me.
Just registered because I wanted to start sharing my experiences using Alexa as a ranking tool, also
because I'd like to hear and maybe share inside info regarding rankings and traffic to compare real info vs. Alexa info.
I've been using Alexa as a tool to track my web site promotion work right from Alexa 1.0, I'd use a simple proportion to calculate and compare traffic. That is: my site has 100 Alexa visits and I know it has 1000 visitors a month (site stats and logs), yours receives 50 Alexa visitors, so I guess you more or less have 500 visitors a month (in the beginning a hypothesis that was full of statistical error, one that I think today is not that flawed due to the large amount of Alexa users, specially on sites that have more than 5000 unique visitors a month).
Sure Alexa is prone to error but considering there are no other options with which you can figure out how much traffic your competition has, Alexa is good enough for me.
You must avoid biasing the info you are trying to observe so be careful NOT to navigate through your sites with the Alexa bar on, only do that the day you check Alexa rankings, you'd affect the ranking too much. That's why I normally use Netscape to surf and IE to do the Alexa stats each time the database updates (each two weeks nowadays, seems). Never tell your clients about Alexa or install the bar on their computers :-) this'll affect the data so much it'll be useless for comparisons.
I am quite sure (based on my site stats) that the lower the ranking, say under 1,000,000, the ranking and traffic are not well correlated, but over that the user sessions stats follow the Alexa ranking pretty well. Furthermore traffic variations are reflected quite well in the ranking too. For example, right now my main site is having less traffic (its a tourist site and the destination is entering low season you'd guess this would follow low season --> less traffic --> lower Alexa ranking, and that is exactly what is happening.
I've filed Alexa info for 14 months now, registering my 40 sites' ranking on each update, plus the ranking of competition sites and things are very clear and the info has been great to stimulate an aggressive promotional job for the sites we manage, if nothing else, Alexa has been a great stimulus for site promotion.
There's a comment from a person having a site that isn't registered or promoted appearing as
#290,000 or something like that. I'd like to ask him, is your site hosted under another domain, as a subfolder of a main domain? Let me clarify, I manage our publicity company's site http://www.domain.net, it's ranking is 110,000 if I build a site under this domain, say http://www.domain.net/mysite/ if I visit my subsite it'll show the ranking of the main site (110000), not the subsite ranking. That is not the case if I create a subdomain mysite.domain.net, then Alexa will show you the ranking of your site and only your site. So if you are building your site on a free hosting site, the ranking you'll see will be the free hosting site's ranking not your site ranking. Don't make that mistake.
Most new sites will not appear on the Alexa database, only once you start promoting them and listing them under search engines and indexes, will any useful ranking appear. If you use domain redirection the real ranking may appear briefly for the redirected domain before Alexa shows the main site ranking. This is the case for a subsite I manage http://www.siteImanage.com, that redirects to http://www.domain.net/siteImanage/
I've just started the promotion of a new hotel site, the first week Alexa would say "no data available" after the first web site promotion session on indexes and search engines and link exchanges with a few strategic sites it appeared
at 3,4 million, last week it was at 2,0 million and surely will be even better on the next update.
What have your experiences been? Anyone out there willing to trade site logs and stats to compare rankings and real life traffic so we can see if reality is followed more or less exactly by Alexa? If so contact me, my highest ranking site is at 43,000 and my lowest around 2,000,000.
Mogens Gallardo Ehlers-Marcussen
I am the person with the unlisted site you wrote of above. No it is not listed as a subdomain. The traffic is all from me. I surf and edit the site wiht the same browser. But I wonder how many users go through the trouble of turning off the Alexa bar when they view their own site? I wish Alexa would have an option to disallow stat counting on ceartin sites for tool bar users. Kinda like IE with the cookie block function where you can always allow cookies from a site or never allow them. If you could put in your own url somehow and that would keep your own visits from skewing the stats for your site that would be great, and make the toolbar much more usefull in my opinion.
your approach of using your own stats, comparing them with Alexa ranking and extrapolating the number of visits of other sites is a nice idea.
You are totally right that it could only be fairly representative if you rank rather high within Alexa and you take care with your own use when visiting your own site. Even so I think their data is so unclear that I doubt it will give you more than an approximation. Also I doubt its accuracy (as they claim themselves) for certain language or country based sites.
For the sake of comparison:
site: 700 unique visitors a working day.
ranking in Alexa: 930.000 ....
funnily enough someone gave it a very good review..
A new search engine powered by Google.
| This 40 message thread spans 2 pages: 40 (  2 ) > > |