Back in December of 2008 I set up a job on one of my servers that once a day searched Google for the title of my book, The Geek Atlas. This job grab the page returned by Google and stored it away.
The first time I ran the search there were 254 pages referencing the term "The Geek Atlas". Most of these were related to Amazon and O'Reilly and were placeholder pages for the book (which wasn't going to come out until June 2009).
I've now processed the data and the chart below shows the number of pages returned by Google per day up to today.
Today there are around 45,000 pages returned by Google. But the curious thing is that the search results show a number of spikes. These spikes are as follows:
1. June 10, 2009 jumped to 82,100 from 7,300 and then back down again the next day
2. June 19, 2009 jumped to 76,500 from 29,600 and then back down again two days later
3. July 13, 2009 jumped to 14,000 from 48,300 and then back down again a day later
So, what happened on those days? The nearest I can come to an explanation is the following:
1. June 10, 2009: Jason Kottke mentioned the book.
2. June 19, 2009: Wired published a full review of the book.
3. July 13, 2009: The BBC published a video and article about the book.
What I'm guessing happens is that the content is syndicated or stolen for use on a variety of short-lived web sites. Hence the spike. If I'm right it tells us that Google is very, very fast at updating its index.