<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	>
<channel>
	<title>Comments on: Alexa</title>
	<atom:link href="http://www.efsavage.com/blog/posts/alexa/feed/" rel="self" type="application/rss+xml" />
	<link>http://www.efsavage.com/blog/posts/alexa/</link>
	<description>Good stuff, updated weekly(ish)</description>
	<pubDate>Wed, 19 Nov 2008 04:56:55 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.6</generator>
		<item>
		<title>By: Yuval</title>
		<link>http://www.efsavage.com/blog/posts/alexa/#comment-4</link>
		<dc:creator>Yuval</dc:creator>
		<pubDate>Tue, 12 Dec 2006 14:12:15 +0000</pubDate>
		<guid isPermaLink="false">http://www.efsavage.com/blog/posts/alexa/#comment-4</guid>
		<description>Hey there,

You are so on the money here about Alexa! I actually wrote a Wall Street Journal reporter who used Alexa on how bogus and stupid it is. If you put a bunch of monkeys in front of machines with Alexa installed, I bet you can sway its 'statistics'. 

Because I work with ad people now, I actually checked with them whether more respected rating services have better techniques. I am talking about Nielsen TNS and comScore. The answer, flatly, is no. EVERYONE of the ranking services uses the same shoddy methodology, using what they see as 'exhibitionist' web users. The reason that they do not follow better techniques, like putting code on the server side (like Google Analytics does) is that they get *demographics* information about the users this way. So people who are on the Nielsen rating system give Nielsen their age, location, favorite color, etc. and in turn they make people believe that they actually vouch for this data. 

This is really a problem...

Yuval</description>
		<content:encoded><![CDATA[<p>Hey there,</p>
<p>You are so on the money here about Alexa! I actually wrote a Wall Street Journal reporter who used Alexa on how bogus and stupid it is. If you put a bunch of monkeys in front of machines with Alexa installed, I bet you can sway its &#8217;statistics&#8217;. </p>
<p>Because I work with ad people now, I actually checked with them whether more respected rating services have better techniques. I am talking about Nielsen TNS and comScore. The answer, flatly, is no. EVERYONE of the ranking services uses the same shoddy methodology, using what they see as &#8216;exhibitionist&#8217; web users. The reason that they do not follow better techniques, like putting code on the server side (like Google Analytics does) is that they get *demographics* information about the users this way. So people who are on the Nielsen rating system give Nielsen their age, location, favorite color, etc. and in turn they make people believe that they actually vouch for this data. </p>
<p>This is really a problem&#8230;</p>
<p>Yuval</p>
]]></content:encoded>
	</item>
</channel>
</rss>
