• Register
• Login

SEO Theory - SEO Theory and Analysis Blog

Algorithm analysis, Web community relationship analysis, SEO practices and techniques, industry news, etc.

• Home
• About SEO Theory
• SEO Glossary

Mismeasuring Market Share: Google’s Twenty Percent

Posted by admin on December 22, 2006 in SEO Metrics, SEO Theory

Xenite.Org gets only about 20% of its visitors from Google. If they banned me today, except for the fact that I do check my rankings, I would barely notice. Nonetheless, there are plenty of people linking to a ridiculous article about Google’s 70% market share. This kind of nonsense is what leads naive Webmasters to foolishly put all their eggs into the Google basket.

What happens to your online business when Google temporarily drops your pages from its listings? Your sales crash through the floor.

Market share is not easy to measure, and you won’t see anything accurate in looking at the top ten or twenty sites reported by large commercial metric services like Hitwise because Hitwise (and its competitors) ignores most Web sites. A metric system is only as good as the data it receives, and the vast majority of Web sites do not carry Hitwise, Google Analytics, Sitemeter, or other third-party metrics.

I’m not talking about mom-and-pop stores, hobbyist Web sites, and other “little corner of the Web” sites. Most Web sites, both commercial and non-commercial, simply do not report their statistics to outside parties capable of aggregating data on a useful scale.

Search engine optimizers are specialists in passing on misinformation. They are nine-tenths propagandists and eight-tenths untrustworthy when it comes to Web traffic statistics and market share. Search engine optimizers, who are expected to deal with statistics and numbers on a daily basis, generally fail to understand the data they see. Pick any big name SEO: Danny Sullivan, Todd Mailcoat, Rand Fishkin, Aaron Wall — any one of them whose good name sends shivers of appreciation throughout the SEO world. You can find bogus numbers and wacky interpretations on all their Web sites.

SEOs love to take incomplete numbers and draw sweeping generalizations from them. And what’s scary is that people pay these guys to bloke smoke out their ears. There is no value in their collective analyses when it comes to determining market share because they don’t have access to, much less use, scientifically valid data samplings. These are third-party pundits who regularly extrapolate from their personal statistics (and, the last time I checked, none of them claimed to cover as many topic areas as I know I do) and the media reports that condense advertising and stat network numbers down to top ten lists.

If we had tried to send men to the moon on this quality of research, John Glenn would have crashed into Mount Olympus on his first attempt to circle the Earth.

When you look at “market share” data, you have to ask some hard questions. For example, if Yahoo! reports 28% and Google reports 45% of all October 2006 quries, how is it that Google sends 6 times as much of all search traffic as Yahoo!?

Search engine optimizers consistently fail to look at all the known facts. One fact they disregard is that Yahoo!’s search results don’t look like Google’s search results. That means a greater percentage of Yahoo!’s referral traffic is going to Web sites other than those appearing in the top placements on Google.

Another fact search engine optimizers conveniently disregard is that most Web sites don’t offer commercial content.

Another fact search engine optimizers disregard is that consumers are increasingly researching products online before making their purchase decisions. So relevant Web site traffic cannot be measured only in terms of conversions and unique site visits. A consumer is likely to look at several sites, often found through more than one search engine, before making a purchase decision.

And what is ironic is that many online marketers have been talking about “the long tail of search” for months. They are aware of the inconsistencies in the reported metrics but several SEO bloggers/journalists have dutifully followed the pack in talking about Google’s “70%” market share.

On the basis of the data directly available to me, Google sends me about 20% of my visitors. We can take several things away from that number. Maybe I don’t have very good placements on Google. After all, I’m only in the top five results for the majority of my more than 100 targeted expressions. Maybe Google would send me 40% of my visitors if I had all number 1 slots.

Maybe I don’t have interesting content. After all, I’m not selling ring-tones and Phentermine, so perhaps I just don’t offer people compelling reasons to visit my pages.

Maybe I should switch statistical packages. After all, my numbers don’t look much like Hitwise’s numbers.

And maybe I’m just doing a better job of marketing my Web sites outside of search engines, so that if for any reason (like a technical glitch) I should temporarily vanish from Google’s search results, I won’t feel the pain so many other people do when their sites vanish.

In this business, it’s actually easier done than said, because it seems like you have to keep saying the obvious over and over again before anyone will listen. Google ain’t everything.

Not yet. Not by a long shot.

Quit paying attention to top ten statistics regardless of how warm and fuzzy you feel about Danny, Rand, and the boys. I read their sites, too. But I don’t follow the pack when it comes to analyzing numbers. I especially don’t give any thought or consideration to “back of the napkin” analysis.