I'm with Rich. I haven't updated in a couple months, but see the search-engines graph in my statistics post. Maybe my readers aren't 100% typical, but they're not that weird.

• Posted by: Tim Bray
• December 19, 2006 8:46 PM

If you add up Google, Ask and AOL (Google & its major distribution partners), I'd agree they're closer to 67-68% than to 60% in the U.S. Certainly, if you look at % of advertiser search spend, they get more 'cause their traffic is worth a bit more (or, rather, Yahoo's distribution partners are worth a lot less).

When you look at Europe, though, Google's at more like 80% and in some countries like Spain and France they're at 95%. Google's essentially a monopoly in Europe.

• Posted by: Chris Zaharias
• December 19, 2006 10:47 PM

Tim's language is interesting. I don't think of the people who visit one page on my site (via search) "my readers." Those who subscribe and read more than a small percentage of what I write are "my readers."

Most of my "visitors" or "searchers" come from Google, as Rich's post suggests. Most of my readers use Bloglines, Google Reader, and a small set of other tools.

• Posted by: Jeremy Zawodny
• December 19, 2006 10:57 PM

I have another theory - Yahoo doesnt provide good search results, and hence user has to search again on yahoo. Therefore yahoo has more "search query" traffic. But since user doesnt click on the result that often, the referrer traffic (our of yahoo) is lower. I think this methodology, of measuring the referrer traffic, correctly reflects the "search" traffic. Yahoo has artifically higher search page views because of their not so good results :)

• Posted by: Analyzer
• December 20, 2006 12:29 AM

The only void in that article is, the method for attaining the stats from the Websites profiled was not revealed. So one can not access the accuracy. However, it appears to be logical.

Would Battellemedia make its stats public just for analysis puposes?

Please consider installing the free version of Sitemeter just for on month.

These are the public stats for the other most popular Technorati blogs

http://seoptimization.blog.com/1221628/

• Posted by: ★ ★ Search Engines WEB ★ ★
• December 20, 2006 1:02 AM

Hitwise has taps on the traffic at a bunch of major ISPs so they're looking at actual web usage traffic on the net. It's not nonsense like 1960's people-meters with ridiculously small sample sizes. Jeremy confirmed the accuracy of the stats inbound to his site as well; this is signficant because, while his blog is popular it's not got youtube's traffic, so the accuracy at the small site level also seems to be quite good.

• Posted by: Rich Skrenta
• December 20, 2006 6:01 AM

For Jeremy Zawodny's site, he is the first result for the name "Jeremy" in Google. Perhaps that is why 90% of his searches come from Google.

• Posted by: Hashim
• December 20, 2006 6:28 AM

Hashim: Good try, but not even close. "Jeremy" is rarely in the top 5 for my site based on search referral logs.

• Posted by: Jeremy Zawodny
• December 20, 2006 6:36 AM

How is the marketshare data collected? I've long suspected that Yahoo's numbers include queries from to their applications (i.e., a search in Yahoo Cars). Although this is indeed a search, the assumption is the query number reflects web searches. Any thoughts on this?

• Posted by: Craig
• December 20, 2006 8:35 AM

How is the marketshare data collected? I've long suspected that Yahoo's numbers include queries from to their applications (i.e., a search in Yahoo Cars). Although this is indeed a search, the assumption is the query number reflects web searches. Any thoughts on this?

• Posted by: arkadaş
• December 20, 2006 8:46 AM

I assume Yahoo's numbers would include that search traffic from their vertical properties. That's entirely appropriate and should be an important source of monetization for them. But they all point at search.yahoo.com, which is where the outbound refer will be seen from.

• Posted by: Rich Skrenta
• December 20, 2006 9:54 AM

For my sites, here are the percentage of search engine-derived traffic over the last two months:

useit.com (content site)
94% Google
4% Yahoo
1% Ask
1% other search engines combined (<0.5%)

nngroup.com (business site)
92% Google
5% Yahoo
1% MSN
2% other search engines combined (<0.5%)

Note that both of these sites are unrepresentative of the broader audience, becuase they target people interested in usability, i.e., members of the high-tech elite.

Also, for these sites, most of the search traffic is irrelevant, consisting of "tourists" and not loyal readers. Users who subscribe to the email newsletter or who go directly to the sites are more important. People who follow links from 3rd-party sites are in-between search users and loyal users (non-search sites are not included in the above statistics, but they account for 35% *more* traffic than search: see previously published chart).

• Posted by: Jakob Nielsen
• December 20, 2006 10:09 AM

Tim Bray, Jeremy Zawodny, and Jakob Nielsen that's some big name commenters!

Heres my two cents...

I think we need to ask these questions:
1. Is the number of search referals a good predictor of number of searches?
For example I use Yahoo to search for podcasts, link analysis etc. But mainly use google to search for websites...
Also as already mentioned here, if one search engine requires more search attempts to find the content, that would skew the results.
However I think the idea that different search engines would be used to find different content is more likely...

2. Do a high percentage of yahoo search queries result in further links to yahoo properties?
If yahoo properties account for 11% of site visits then conceivably a lot of there search traffic is going to another yahoo property. Yahoo properties would push you towards using the yahoo search then push you back towards more yahoo stuff...

3. Just because Hitwise provides accurate statistics for topix.net does that mean it is accurate for other sites?

If I understand Rich Skrenta he has looked at the stats for his site topix.net compared it with hitwise then effectively extrapolated that the same relationship will hold for other sites. This may not be true...
Remember hitwise are the same guys who are reporting a 60% google market share...

Out of interest over the last month my site (http://www.jambe.co.nz) seems to have had 58% google 20% yahoo and the other 20% from a variety of very small numbers. This is a very low traffic site so I wouldn't put much weight on it...
I don't have the stats for my blog at the moment, I have hittail running but it doesn't give you counts of total number of search referrals by searchengines.

• Posted by: James Brunskill
• December 20, 2006 2:01 PM

Google crawls and indexes faster than the others, so for very dynamic sites they will deliver more traffic. Topix.net is definitely a more dynamic site. For www.searchles.com we get about 95% of our search engine traffic from Google because our content is quite amorphous.

• Posted by: dumbfounder
• December 20, 2006 2:46 PM