This other article on HN shows that in Q4 2007, there were large spikes in seemingly normal search terms: http://nikgregory.com/2010/11/bizarre-google-trend-sea

This other article on HN shows that in Q4 2007, there were large spikes in seemingly normal search terms: http://nikgregory.com/2010/11/bizarre-google-trend-search-spike/ http://nikgregory.com/2010/11/bizarre-google-trend-search-sp... I wouldn't read too much into Google's protection against search trends. Google cares about its search algorithm, but probably not as much about the trends. Clearly if you do your programming and distribution correctly, Google shouldn't be able to figure out that it was calculated. If anything, they should look at how organic the follow-through on those searches are: If the user searches for something, do they actually click on it or was it just a search? But then again, checking if the user follows through technically shouldn't factor in, after all, how many people click through to something after searching for "Weather New York, NY"? (Although they probably check if at the very least the HTML was displayed, using their JavaScript). Getting legitimate, unconnected searches is enough effort. Unless you choose a quite obscure search term, you should need quite a large volume in order to game the system in some sort of profitable/notable way.