The “long tail“, in addition to being the title of one of my favorite books (and a must-read for anyone engaged in business online), is an oft-referenced term in our industry. In the world of search engine marketing, the long tail refers to the large number of search keywords that are long or specific, and searched not often. For example, “los angeles dentist” is a commonly searched keyword, while “where can i find a dentist in los angeles who works on kids” is very much a long tail keyword.
Anyhow, in one session at 2007’s Pubcon, I heard a statistic that made me take note: that upwards of 40% of all queries that Google sees every single day, it has never seen before.
I scoured for verification of this statistic, and while I did not find the 40% number repeated, I did find explicit mention that 20-25% of Google queries in 2006 it had never seen before. Perhaps the 40% is bloated, or perhaps the percentage has actually risen from 25% to 40% in the year or so since the last statistic. To me, however, the implications of any of these numbers are vast.
This means that of the tens of millions of searches performed each day, between 20-40% of them Google has never seen searched before, and will most likely never see again, otherwise their search volume would be expanding exponentially (it’s fast, but not that fast).
This is a difficult concept to get one’s head around. I can explain some of these unique queries to items in the news searched one day and not again, spelling errors that are bizarre and unlikely to be repeated, strange research-based queries and full sentences and questions entered as search queries. Those scenarios out of the way, it is still hard for me to fathom the sheer volume of unique queries that Google is seeing and processing, one at a time. It makes you realize what Google’s Udi Manber said quite eloquently: Search is a hard problem.
While I can neither explain or prove this unique query statistic, I can at least see if I am “part of the problem.” I am going to make a conscious effort for the next few days to note the queries I make in Google, and at the end try to evaluate the likelihood that that is a truly unique query, at least in my world of searching.
And no, I won’t pad the stats by smashing my hand on the keyboard and hitting submit!