Comment Not a user-centric way of measuring "success rate" (Score 1) 385
The definition of success-rate is not user-centric.
'The market watcher defines "success rate" as the percentage of search queries that result in a visit to a website.'
How do we know this matches what a user finds interesting? As someone mentioned, typically with Google you don't even need to go to a website, the answer you're looking for is already shown in the search results. That may not sit well with the website owner, and this study is clearly measuring success from the perspective from the website owner.
But who cares about that when you want to find something? Clearly the users are voting with their feet.
Another interesting thing: what if, after a Google search, you're happy with the first search result. That's one hit. But if you go to Bing, say the first result is not relevant and you need to visit the first 5 links. Will that make Bing 5 times as "successful"?
Certainly from the perspective of the website owner or advertiser, but surely not from the user.
So I don't think the way Experian Hitwise measures has a lot of bearing on "success" from the perspective of a user. And that is what makes things popular, not whether it satisfies the advertiser.