I highly doubt their tactics here. They're grabbing aggregate ISP data for users (http://www.hitwise.com/us/about-us/how-we-do-it) and running scripts to see if the search is followed by an address that appears on the search page. Also, it seems like they don't account for google's instant search, but it's hard to tell as their first "success" statistic was posted in October (Instant Search was released on September 8th). They do however, publish upstream traffic percentages for sites in various categories (http://www.hitwise.com/us/press-center/press-releases/google...) prior to September. Note that in this case, the ratio between the upstream click throughs (75% for google, 15% for yahoo, and 10% for bing for health related click throughs) and the ratios of the aggregate search traffic (71.4% for google, 15% for yahoo, and 9.4% for bing during that month) allow us to extrapolate the "success" rates during this period, and it seems to suggest that everyone is in line for search results (with google being slightly ahead, at least for health. I didn't run the numbers for the other categories as we dont know the search volume for each of those, and as such you couldn't find an average statistic that was relevant). Just my two cents, but it seems like the data is highly affected by google's instant search.