Talk:Alexa Internet/Archives/2012

minority?

"Mozilla Firefox, Opera, or other minority web browsers." does anybody else find a problem with this? JohnRussell 00:28, 20 January 2006 (UTC)

Not really, I use Firefox, but I admit that there's a huge silent majority who rely on Internet explorer. However, the sentence could be construed by the overly PC-conscious user as meaning that firefox/opera are used solely by ethnic minorities, so a reword couldnt hurt. GeeJo _{(t) (c)} • 23:08, 21 January 2006 (UTC)

It's not just Firefox it doesn't work on. It's amost every computer, mobile phone and PDA except Windows/IE5+. I've made this clearer in the article. PS Sorry about the accidental (m) tag. Stephen B Streater 08:39, 21 March 2006 (UTC)

The information about Alexa support of other browsers is very misleading at its website. One place contradicts another. In order to support representativeness of the sample Alexa says one thing: "Alexa's sample includes users of Internet Explorer, Firefox and Mozilla browsers." and "Alexa sample includes toolbars built for Windows, Macintosh and Linux." [1]. However, when I tried to download a toolbar for linux/firefox, Alexa's website tells quite a different story and encourages me to create one myself [2]. One question is unclear: if somebody creates this toolbar, will Alexa count visits generated by this toolbar or not? IMHO, Alexa should disclose the sample size on which its ratings are based. Howard Metzenberg estimated this sample to be around 180,000 in 2003 [3], but now it seems not possible to reestimate using the same method, since Alexa no longer shoing the daily data of low traffic websites on which that estimation was based. Information about number of downloads of the toolbar really tells nothing. Alex Kosorukoff 03:45, 12 June 2006 (UTC)

You could ask Alexa how they count. My understanding of their methodology is that any application that uses their service gets counted. Essentially, related links is a web service, and traffic stats are computed from an analysis of this service's web logs. So if you created a Linux Alexa client, it could contribute to Alexa's stats.

That said, I'm only guessing based on my familiarity with Alexa's stats circa 2000. The architecture and counting methods could be different now, and the best way to get the answer to your question is to ask them. --Zippy 18:41, 3 November 2006 (UTC)

Web traffic

How does it measure how much "web traffic" a site receives? - Jerryseinfeld 00:01, 14 Nov 2004 (UTC)

Alexa collects statistics on the behavior of web surfers who use the Alexa toolbar and related services. It uses this sample of internet users to generate site statistics. I believe they describe this on their site. --Zippy 05:56, 30 September 2005 (UTC)

Are the Alexa traffic rankings determiend solely by this toolbar (this is all I've read). This seems like it would lead to a huge selection sampling bias in the stats produced by this. The traffic stats would be those for user who (a) know about Alexa and (b) choose to install the toolbar. Dharris 02:18, 3 October 2005 (UTC)

I've just researched this bias question a little further and found that Alexa does indeed have a page about biases which lists, among others, biases about the toolbar being for IE only, Windows only, and having a bias towards Korean sites. Dharris 02:25, 3 October 2005 (UTC)

They also collect data from third-party tools like SearchStatus. —The preceding unsigned comment was added by 67.164.56.2 (talk) 04:09, 25 January 2007 (UTC).

Is that what claiming that they now take into account more data sources "beyond Alexa Toolbar users" means? The link after that claim is dead. It should be clarified that those "more data sources" are just more Alexa toolbars that not many people use. Still it's strange that SearchStatus and Sparky got accepted by addons.mozilla.org if they're in fact spyware.--87.162.27.241 (talk) 13:11, 23 April 2009 (UTC)

Poor article

I've read this article twice and still don't know what Alexa is. This is a really poor article, from structure to everything else. I wish I could fix it but I don't know what Alexa is yet. —Cantus…☎ 04:04, Feb 7, 2005 (UTC)

One problem is that Alexa has changed greatly over the years, going from a startup whose focus was to provide related links (that is, you're looking at a web page, what else might you want to see) to its acquisition by Amazon, where for a year it was focused on creating a comparison shopping service called zBubbles (which only barely went live, in the sense that the service was barely promoted), and since then, has largely been known for collecting and publishing statistics on the popularity of websites. In this role, many webmasters use Alexa's free stats as a way to tell how they are doing vs competing sites.

I agree, the article does need a major rewrite. I'd be happy to fill in details about the company if needed -- I was there before and after the Amazon acquisition. --Zippy 05:54, 30 September 2005 (UTC)

I agree completely with you. I was extremely puzzled, I had a very very vague notion after reading the article. I'm all for merging these two articles, I think that Global 500 could be a sub-heading for Alexa, seeing as they are so closely related, and the Global 500 Article is only a sentence or two. —The preceding unsigned comment was added by 72.138.194.62 (talk) 19:35, 18 February 2007 (UTC).

What Alexa Is

I believe Alexa bills itself as the "Web Information Company". The Alexa toolbar tries to make your browsing easier by providing information on "Where you are" and "Where to go next". The Alexa website give the same information as the toolbar, but in greater detail and without requiring you to install a toolbar. More details on Alexa are available at http://pages.alexa.com/company/index.html and at http://pages.alexa.com/company/history.html These pages give a much better picture than this article. The way Alexa determines its web traffic numbers is described at http://pages.alexa.com/prod_serv/traffic_learn_more.html

I have some experience with Alexa's numbers. I wrote the first scripts Alexa used to generate site traffic reports. I suspect the methodology and sampling issues are similar, if not the same, as what I first did for Alexa in 1998. Note, I'm not offering that we put original research into the article, only that I am willing to help interpret primary documentation on how the service works for the purpose of this article.

One point I note is that, while I was there, Alexa was not IE only, but was also built into Netscape as the "What's Related" feature. We also offered a bookmarklet version of the tool for AOL and Mac users. Prior to the bookmarklets, there was a Mac toolbar as well. --Zippy 02:09, 6 March 2006 (UTC)

WWW vs. St.

Any discussion of whether an off-site link should be included prominently that doesn't at least pay passing mention to the 65 character's worth of Wikipedia's attention to the street address of the company (in the first faqqin paragraph no less, which is great) that were previously deemed appropriate, and were also left by the most recent reverter/updater will be a fun discussion in which to participate.

As mentioned in the subject of my initial sumission, a primary motivation was the inappropriateness of the word "website" being made to be a link to the Wikipedia article on website, at least as it was in the then current version of the article. It just didn't fit with current norms or current online grammer. Simply put, if you're gonna allow an article about something that is, essentially, a web site.

Additionally, if somewhat disconnectedly: if you're going to furthermore ordain that if someone who comes to wikipedia.org and searches for "alexa.com" is then REDIRECTED without explanation to an entry that doesn't include "alexa.com", then perhaps the prominent inclusion of the site in question is in order. -:)Ozzyslovechild 02:15, 3 January 2006 (UTC)

confusing definitions on alexa.com

hi everyone. have you ever typed the name of a website and read the details about how alexa does its traffic ranking? i find it really confusing. for exemple in the part called "what is reach?", they talk both about percentages and million users. let's take wikipedia as an exemple. ;-) if the figure "reach per million users" is written 30,860, does it mean that 30,860 million users went on it? or is it the percentage of a million of users. so only around 30,860 people went? or 30,860 people in a million..which means that on a population of 10 million people you multiply it by 10? why then choose such a weird way of calculating? instead of an easy percentage or total number of users?!!! seriously..how can such a confusing method based on fancy way to rank the traffic be famous and successful... thanks for your help if you can give more explanation about it. Manzanita.

It's the number of people who watch it out of every million who use the internet (and run alexa). The reason they don't use percentages is that some sites have 1 person in a million watching - 0.0001% - and this is a bit small as a percentage. For example, my company's site, Forbidden Technologies, has a few people out of every million watching it. % is per hundred - and for rarer events, people use per thousand or per million or per billion. Stephen B Streater 21:18, 28 March 2006 (UTC)

ok thank you Stephen, i better understand. so for exemple 30,860 "reach per million users" means that as alexa is supposed to have 10 million people using its toolbar, then it means that basically 308 600 people of these 10 million using the toolbar go on wikipedia. which is anyway 3% of those users. i just still find that their explanation in "what's this?" is confusing..ranking.com looks for exemple clearer, even if their sample of people is only of 200,000 people using their BrowserAccelerator toolbar. thanks anyway. Manzanita

You're welcome. Stephen B Streater 15:42, 29 March 2006 (UTC)

Calculating the approximate number of users of a given site from the Alexa reach

Is it correct to multiply the Alexa reach (the users of a given site out of a million internet users) by the approximate total number of millions of internet users (as given in Global internet usage), that is, as of March 2006, (Alexa reach * 1023) ? If so, maybe this piece of info might make the article clearer.

No, because of the considerable sampling biases of Alexa, especially the bias excluding non English speaking users. David.Monniaux 18:17, 12 November 2006 (UTC)

article name

What was wrong with "Alexa Internet"; that name appears on their current website. Failing that, I think the article should be renamed to Alexa (internet). John Vandenberg 09:12, 13 January 2007 (UTC) Contributions/91.110.158.116 (talk) 19:28, 15 June 2009 (UTC) The Alexa Article could also be called Alexa(website)

The name of the company is "Alexa Internet", not just "Alexa". —The preceding unsigned comment was added by 67.164.56.2 (talk) 04:11, 25 January 2007 (UTC).

Alexa score

I was just reading the Kurnik article and came across this sentence: "The Polish version has an Alexa Internet score of 6,707". What does the score signify? This article has nothing about it. --ZeroOne (talk | @) 20:45, 14 January 2007 (UTC)

I've changed this to say something I understand. Stephen B Streater 04:03, 26 January 2007 (UTC)

merge from Global 500

The "Global 500" article appears to refer to this Alexa page which does not use the term "Global 500" but instead is simply entitled "Top Sites". Shawnc 16:48, 20 January 2007 (UTC)

The article title was changed to "Alexa (Internet)" probably to indicate that it's Alexa as related to the internet. It should be changed back to "Alexa Internet" because that's the full name of the company, not a description of the usage domain.

Alexa links not allowed?

I noticed here that Shadowbot removed a link referencing the traffic a site gets, according to Alexa. Is there some policy about Alexa links that I haven't read? Blast 11.04.07 2206 (UTC)

Nope, no policy. If you disagree with a bot change, just revert it. --- RockMFR 22:26, 11 April 2007 (UTC)

Alexa vs Internet Explorer

If you were to run anti-spyware on a newly installed Windows machine with Internet Explorer 5 or 6 freshly installed, does anyone else come up with Alexa as pre-installed? I always find it is installed, and AdAware detects it. I've yet to have proof this is a normal thing, but if so, would it be wise adding this to the article? QuickHare 22:27, 2 July 2007 (UTC)

In Internet Explorer 5 and 6 there was a "Related Links" feature that used to show similar websites to the one you are on accessible on the Tools menu. This launched a sidebar and the results came back from Alexa. I'm surprised the main article doesn't mention this. Removed in the SP2 update for IE6 as part of XP SP2 I think, along with the Media sidepanel that integrated WMP into IE. 86.167.100.78 (talk) 22:27, 19 November 2008 (UTC)

Brewster Kahle and Alexa

I noticed that Brewster Kahle was listed in the key people field on Alexa's infobox. I removed him since he's no longer at Alexa. According to Alexa's management page Bruce Gilliat is now the CEO. --Zippy 20:42, 26 September 2007 (UTC)

Wikiscanner question

8 edits on this page can be traced back to an Alexa office in San Francisco. [4] Of course, that same IP range is responsible for editing the Gilmore Girls page [5] , so it might be a customer and not an employee (is that really what goes on at their office?). Is self-editing allowable under Wikipedia guidelines? —Preceding unsigned comment added by Notostraca (talk • contribs) 18:56, 5 November 2007 (UTC)

The link cited to scrutinize Alexa edits is now found through [6] . Piano non troppo (talk) 15:33, 17 April 2009 (UTC)

The exact content of those edits I haven't reviewed yet, though. Oh, and that previous edit was me. Notostraca 19:05, 5 November 2007 (UTC)

There's nothing to stop Editors working on their own pages, but it does cause problems. (Here in the UK we have caught out political parties making changes to suit their leader's latest pronouncements!).

This was well worth pointing out, and I hope the Alexa employee who is also a fan of the Gilmore Girls has the sense to leave well alone. If you can tell when they have made changes it's worth scrutinising them for bias and alerting others if you detect it. OTOH of course, they may just be making changes for the sake of accuracy, in which case - well done!  :) raining girl (talk) 10:30, 19 May 2009 (UTC)

Bias?

The second two paragraphs out of the four in the "concerns" section seem like they are from a biased, not a neutral, point of view. Regardless of whether they are biased or not, none of the claims they make are substantiated by any means, and they are poorly written. ChicagoGooner (talk) 04:57, 20 November 2007 (UTC)

Wrong external links

I discovered that all the external links in the "See also" section refer to "http://www.alexarankteam.com/" despite they are associated with the competitors names of Alexa.

I've changed the links into the right ones, as in the french page of Alexa Internet. —Preceding unsigned comment added by 90.18.149.63 (talk) 18:12, 5 March 2008 (UTC)

Apparently the same false link has been used for the "External links" section, this time i have managed to undo the modification (yes, i'm a begginer in Wikipedia editing !).

It's the same IP for these 2 cases of spamming : 64.15.147.240. I suggest to block this user. —Preceding unsigned comment added by 90.18.149.63 (talk) 18:43, 5 March 2008 (UTC)

hay, actually a discussion about this. It was wrong and and I fixed it, hope it stays fixed. —Preceding unsigned comment added by 24.9.82.19 (talk) 06:42, 13 February 2009 (UTC)

Description of page rank?

It's not clear either from this article or from anything I can find at Alexa.com how to actually interpret their page rank. Is a lower number like one a higher rank (that's what I would think, but still I wonder), or would a high number like 1,000,000 be a higher rank? Can this be clearly stated in the article? Anyone know a page at Alexa.com that actually says anything about it?--Narfnarfsillywilly (talk) 20:59, 24 July 2008 (UTC)

A lower number is a better rank. I'm pretty sure it's explained on their site somewhere. --- RockMFR 00:57, 25 July 2008 (UTC)

Template?

What about a navigational template of some of the top sites by Alexa rank, say the top 10 or 20? Richard001 (talk) 10:25, 26 October 2008 (UTC)

Alexa's own rank

Just a passing by question: Alexa doesn't rank itself, but where would it fall if it did? Are there similar sites we can use to get a rank for it? Richard001 (talk) 06:52, 8 December 2008 (UTC)

Redesign and New Stats

I removed some words that implied a relationship between the outage and the site redesign. Though they may very well be related, that fact is stated in neither of the site notes, and has the appearance of original research. 98.150.85.243 (talk) 01:16, 20 April 2009 (UTC)

Quality Standards Re-writes

Since the template was stuck at the top of this article mentioning Quality Standards, several editors (including myself) have made substantial changes. I've focused most on bias and the removal of non-encyclopaedic language. Others have added information and clarified points.

Can some more experienced Quality Standards Gonk please take a look and make suggestions as to where to go next, or remove the flag, please? Thanks. raining girl (talk) 10:26, 19 May 2009 (UTC)

OR, I know but even so

We have two largish websites the Schools Wikipedia (12000+ UIPs a day, online) which shows a rank of 240906 and www.soschildrensvillages.org.uk (3500 USPs a day) a rank of 198,243. The UIPs are done on the same stats package.

We know from the usage pattern that the Schools Wikipedia is heavily used by schools who presumably do not have Alexa toolbars fitted (most schools use the off line version and are not in the stats anyway) but even so the fact the sites are not even ordered correctly is rather striking. --BozMo talk 13:54, 1 October 2009 (UTC)

Products

We need a section on that. From the history it's not really clear what those were/are. Were the Alexa Image Search and Alexa Web Search different products or did they tie in with the toolbar/traffic business? Pcap ping 13:47, 4 January 2010 (UTC)

Peter Norvig's examples

This revert claims it's a WP:SYNT to include them, but that's clearly false because the main point is advanced by the two books cited. Norvig's data is just an example of that, and clearly written as such. He is a distinguished computer scientist, so a reliable enough source for something like a data point (per WP:RS). Besides, the information is clearly attributed to him (per WP:NPOV). Pcap ping 23:55, 4 January 2010 (UTC)

The main problem I see is that the paragraph is generally trying to advance the position that Alexa sucks at low-traffic websites stats. The Norvig article does not state this; rather, it gives some traffic data on a handful of websites that just happen to be low-traffic. But Norvig never says anything about these sites being low traffic, nor does he arrive at the conclusion that the 50:1 difference is due to low-traffic and/or sample sizes. Furthermore, I don't think an experiment with five data points can be considered meaningful or scientific. And the writer of the article does not appear to be an expert statistician, so I don't believe it is appropriate to use a self-published article from him which is primarily about statistics. — RockMFR 00:22, 5 January 2010 (UTC)

Alexa website traffic reporting RS/N

WP:RS issues related to Alexa and website traffic reporting in general are currently being considered in an RS/N. Opinions/observations of any interested editors are solicited and welcomed. JakeInJoisey (talk) 12:23, 23 April 2010 (UTC)

Rewrite request - criticisms section

"Other tools like Google toolbar also send the potentially private data after user approval."

What is "potentially private data?" Could someone who knows what this is trying to say rewrite it please. —Preceding unsigned comment added by 81.108.241.233 (talk) 15:49, 18 November 2010 (UTC)

Alexa Unique visitors and relibility of Alexa rating

I think we need information about formulas for translation Alexa rating to unique visitos per day. Also we need researches about reliability of Alexa rating.

Example of this material. Compare Alexa with another ratings. Unique vistors metrics. Reliability of data

Use Google Trans for reading. Another sources with unique visitors/reliability research? --NoPR (talk) 17:16, 19 August 2011 (UTC)