When I posted Enquisite’s
search engine popularity stats yesterday, I almost went into a riff about how site
owner traffic stats are often so different from what the major measuring services
report. Google almost always comes out much, much higher when you look at site
traffic referral sources (as with Enquisite’s stats). I did get into this on
SearchCast. But Rich Skrenta does a great job on the issue for who prefer to
read about the issue. More on his post, along with a long look at stats, the
pros and cons of Google being a huge traffic driver, and more.
Google’s true search market share is 70%is Rich’s article. He highlights how
the ratings services tend to put Google in the 40 percent range, but site owner
traffic shows Google with a much higher than that. Having access to Hitwise’s stats, he
fires up a chart showing the traffic various sites get from different search
engines — Apple, Craigslist, eBay and others. The percentage of Google traffic
far outweighs that from other sites.
To be super clear, let’s zoom in on one site, Apple. Hitwise reported that
8.6 percent of its search related traffic came from Google, with Yahoo behind at
Now if you work off the general benchmark that Google is reported to have 40
percent of the search share, versus Yahoo generally around 20 percent, you can see the
gap. The percent of traffic Yahoo sends should be about half the 8.6 percent
Google sends — or 4.3 percent. It’s not. It’s that lower 2.4 percent figure.
Of course, "general benchmark" is not entirely accurate. To explain more, I
have to go back to that
Search Popularity Stats, Sliced & Diced article I did
last month. Here were the figures for all search engines, for traffic in the US,
for October 2006:
And here’s a zoom in on just Google from October 2005 through October 2006:
On average (comScore, NetRatings & Hitwise for October 2006), Google has a 52
percent share. That’s higher than the "40-something percent market share" that
Rich talks about. For ages, 40-something would have been fine to say, but
50-something is more accurate. But he’s still dead on target that if you talk to
site owners, Google drives a far higher percentage than the share of searches
happening. It’s more like 70 to 90-something for them.
Remember my Search
Engine Land Stats: December 2006 Preview that I did last week? That
illustrates the case so well. We had 800 search related visits from Google for
the first week or so in December, compared to 37 (37!) from Yahoo, the next
highest search engine.
Oh, but Search Engine Land is such a new site. Sure, but that shouldn’t have
so much to do with it. Yahoo’s supposed to be a good search engine. It should be
able to find all of our pages just like Google. And, I can
see we have plenty listed. Those pages so far just haven’t sent us much
Well, maybe we haven’t optimized right for Yahoo. Well, we haven’t optimized
for Google. We’ve optimized generally, doing the things any search engine wants.
But Google delivers more.
Just a fluke for Search Engine Land? No — at Search Engine Watch, Google was
also the biggest driver. How about my personal blog,
Daggle? For Dec 1 through 19, I had 18,098
"organic" or search-related visits. Google sent 16,839 of those — 93 percent.
MSN was next highest at 383 or 2.1 percent, followed by Yahoo with 349 or 1.9
Ask any site owner, and almost consistently, they tell you Google is by far
the biggest search engine source they have. That makes life pretty sucky for
Consider the routine freak-outs that happen if Google makes an algorithm
change and someone loses traffic. They get blamed for hating webmasters; for
trying to drive up AdWords purchases; for trying to promote their own content;
for dissing the little guy. You name it. Heck, you’ve even got some people like
who think they shouldn’t do algorithm changes during the holidays to
especially protect site owners from any harm.
If Yahoo or Microsoft has a hiccup, most site owners don’t care. They don’t
care because they don’t notice. They don’t notice because they aren’t getting
much traffic relatively speaking. And if they aren’t getting much traffic, then
where are all those searchers going if comScore or NetRatings or Hitwise say the
search stats are higher.
They might not be going to web search. Lots of stuff is mixed in with search
data. In the New Year, I’ll try to do an entire new drill down into the stats.
But a search for something on Yahoo Sports? That might be counted as a "search,"
and it is — but it’s not the type of search that would register with site-based
metrics. The searcher might stay entirely inside Yahoo.
Back to Google, I feel for them (as I talked about more on yesterday’s Daily
SearchCast). The clearly send so much traffic — and traffic for free — to
websites and get so much grief in return because of it. If they were less
dominant, the Blame Google mentality (and don’t bother checking anyone else)
would diminish. But then so would the traffic site owners love from Google.
That’s one reason why I love what Google’s been doing with
Google Webmaster Central. I
talked about this in my
keynote at SES Chicago this month, plus
it was on
my 25 Things I Love About Google list earlier this year. Big bad evil Google, as
it is often painted, has consistently been adding more tools and offering more
advice to support site owners who want the free traffic from Google. Yes,
there’s a good PR aspect to doing this. But there’s also a sense of relief for
me that Google — so fixated on chasing the video dollar, the radio buck, the
newspaper cash, throwing ads up wherever we are — actually is putting in
significant time and resources to help site owners.
I’ll leave off with some reading materials on stats, if you want to get into
them more, from my past writings over at Search Engine Watch:
Figures Show First Google Decline For Nearly A Year, But What To Believe?:
First in a series I did earlier this year, it looks at many of the issues in
knowing what goes into some of the figures we get from ratings services.
NetRatings: Tiny Google Decline, But Not The First Time & Yahoo Growth:
Second in my series, the focus shifted to stats from NetRatings.
Google & Yahoo Make Tiny Gains In July 2006: Third in the series, the
focus was on Hitwise.
Popularity Stats, Sliced & Diced: Fourth in the series, it took stats
from all the services over time and charted them out, plus offered tips on
understanding why they don’t all agree.
Ain’t So Bad, Says Site Search Comparison: Alexa figures depend on
data sent back by the Alexa toolbar, and many feel these can easily be gamed.
Links to background on that, plus some support for Alexa.
Don’t External Site Popularity Estimates Add Up?: Covers both a good
BusinessWeek article (Web
Numbers: What’s Real?) on why your own internal metrics might not match
ratings services plus
Analytics vs. Competitive Intelligence Metrics from SEOmoz, which shows
among other things, a number of sites listing Google as a top referring
WebSideStory: Google Continues Its Popularity Rise: Unlike comScore,
NetRatings and Hitwise, which take in data based from monitoring users, doing
surveys or using ISP traffic information, WebSideStory data comes from traffic
hitting actual web sites. This shows you the Google climb over time. Back when
the others were putting Google in the high 40 percent share range,
WebSideStory was in the mid-50 percent. Unfortunately, the company doesn’t
consistently issue stats on a regular basis (press release page is
Says Google Stays On Top In Global Search: OneStat is similar to
WebSideStory, depending on site owner traffic data to do ratings. The downside
to traffic data is that there are other types of skews they can have, such as
the Yahoo Sports example above. That is a search — and it is a share of
searches that should be measured. It just might not translate into site owner